Line data Source code
1 : !--------------------------------------------------------------------------------------------------!
2 : ! CP2K: A general program to perform molecular dynamics simulations !
3 : ! Copyright 2000-2025 CP2K developers group <https://cp2k.org> !
4 : ! !
5 : ! SPDX-License-Identifier: GPL-2.0-or-later !
6 : !--------------------------------------------------------------------------------------------------!
7 :
8 : ! **************************************************************************************************
9 : !> \brief input section for MP2
10 : !> \par History
11 : !> 05.2011 created
12 : !> \author MDB
13 : ! **************************************************************************************************
14 : MODULE input_cp2k_mp2
15 : USE bibliography, ONLY: &
16 : Bates2013, DelBen2012, DelBen2013, DelBen2015, DelBen2015b, Rybkin2016, Wilhelm2016a, &
17 : Wilhelm2016b, Wilhelm2017, Wilhelm2018, Stein2022, Stein2024, Bussy2023
18 : USE cp_eri_mme_interface, ONLY: create_eri_mme_section
19 : USE cp_output_handling, ONLY: add_last_numeric, &
20 : cp_print_key_section_create, &
21 : debug_print_level, &
22 : high_print_level, &
23 : low_print_level, &
24 : medium_print_level, &
25 : silent_print_level
26 : USE cp_units, ONLY: cp_unit_to_cp2k
27 : USE input_constants, ONLY: &
28 : bse_fulldiag, bse_iterdiag, bse_tda, bse_abba, bse_both, &
29 : bse_screening_w0, bse_screening_tdhf, bse_screening_rpa, bse_screening_alpha, &
30 : bse_iter_both_cond, bse_iter_en_cond, bse_iter_res_cond, bse_singlet, &
31 : bse_triplet, do_eri_gpw, do_eri_mme, do_eri_os, do_potential_coulomb, do_potential_id, &
32 : do_potential_long, do_potential_mix_cl, do_potential_short, do_potential_truncated, &
33 : do_potential_tshpsc, eri_default, gaussian, gw_no_print_exx, gw_pade_approx, gw_print_exx, &
34 : gw_read_exx, gw_skip_for_regtest, gw_two_pole_model, kp_weights_W_auto, &
35 : kp_weights_W_tailored, kp_weights_W_uniform, mp2_method_direct, mp2_method_gpw, &
36 : mp2_method_none, numerical, ot_precond_full_all, ot_precond_full_kinetic, &
37 : ot_precond_full_single, ot_precond_full_single_inverse, ot_precond_none, &
38 : ot_precond_s_inverse, ri_default, ri_rpa_g0w0_crossing_bisection, &
39 : ri_rpa_g0w0_crossing_newton, ri_rpa_g0w0_crossing_z_shot, soc_lda, soc_none, soc_pbe, &
40 : wfc_mm_style_gemm, wfc_mm_style_syrk, z_solver_cg, z_solver_pople, z_solver_richardson, &
41 : z_solver_sd, rpa_exchange_none, rpa_exchange_axk, rpa_exchange_sosex, G0W0, evGW0, evGW, &
42 : sigma_none, sigma_PBE0_S1, sigma_PBE0_S2, sigma_PBE_S1, sigma_PBE_S2
43 : USE input_cp2k_hfx, ONLY: create_hfx_section
44 : USE input_cp2k_kpoints, ONLY: create_kpoint_set_section
45 : USE input_keyword_types, ONLY: keyword_create, &
46 : keyword_release, &
47 : keyword_type
48 : USE input_section_types, ONLY: section_add_keyword, &
49 : section_add_subsection, &
50 : section_create, &
51 : section_release, &
52 : section_type
53 :
54 : USE input_val_types, ONLY: char_t, &
55 : integer_t, &
56 : logical_t, &
57 : real_t
58 : USE kinds, ONLY: dp
59 : USE string_utilities, ONLY: newline, &
60 : s2a
61 : #include "./base/base_uses.f90"
62 :
63 : IMPLICIT NONE
64 : PRIVATE
65 :
66 : CHARACTER(len=*), PARAMETER, PRIVATE :: moduleN = 'input_cp2k_mp2'
67 :
68 : PUBLIC :: create_mp2_section
69 :
70 : CONTAINS
71 :
72 : ! **************************************************************************************************
73 : !> \brief creates the input section for the mp2 part
74 : !> \param section the section to create
75 : !> \author MDB
76 : ! **************************************************************************************************
77 64714 : SUBROUTINE create_mp2_section(section)
78 : TYPE(section_type), POINTER :: section
79 :
80 : TYPE(keyword_type), POINTER :: keyword
81 : TYPE(section_type), POINTER :: print_key, subsection
82 :
83 64714 : CPASSERT(.NOT. ASSOCIATED(section))
84 : CALL section_create(section, __LOCATION__, name="WF_CORRELATION", &
85 : description="Sets up the wavefunction-based correlation methods as MP2, "// &
86 : "RI-MP2, RI-SOS-MP2, RI-RPA and GW (inside RI-RPA). ", &
87 : n_keywords=4, n_subsections=7, repeats=.TRUE., &
88 : citations=(/DelBen2012, DelBen2013, DelBen2015, DelBen2015b, Rybkin2016, &
89 : Wilhelm2016a, Wilhelm2016b, Wilhelm2017, Wilhelm2018, Stein2022, &
90 841282 : Stein2024, Bussy2023/))
91 :
92 64714 : NULLIFY (keyword, subsection)
93 :
94 : CALL keyword_create( &
95 : keyword, __LOCATION__, &
96 : name="MEMORY", &
97 : description="Maximum allowed total memory usage during MP2 methods [MiB].", &
98 : usage="MEMORY 1500 ", &
99 64714 : default_r_val=1.024E+3_dp)
100 64714 : CALL section_add_keyword(section, keyword)
101 64714 : CALL keyword_release(keyword)
102 :
103 : CALL keyword_create( &
104 : keyword, __LOCATION__, &
105 : name="E_GAP", &
106 : description="Gap energy for integration grids in Hartree. Defaults to -1.0 (automatic determination). "// &
107 : "Recommended to set if several RPA or SOS-MP2 gradient calculations are requested or to be restarted. "// &
108 : "In this way, differences of integration grids across different runs are removed as CP2K "// &
109 : "does not include derivatives thereof.", &
110 : usage="E_GAP 0.5", &
111 64714 : default_r_val=-1.0_dp)
112 64714 : CALL section_add_keyword(section, keyword)
113 64714 : CALL keyword_release(keyword)
114 :
115 : CALL keyword_create( &
116 : keyword, __LOCATION__, &
117 : name="E_RANGE", &
118 : description="Energy range (ratio of largest and smallest) energy difference "// &
119 : "of unoccupied and occupied orbitals for integration grids. Defaults to 0.0 (automatic determination). "// &
120 : "Recommended to set if several RPA or SOS-MP2 gradient calculations are requested or to be restarted. "// &
121 : "In this way, differences of integration grids across different runs are removed as CP2K "// &
122 : "does not include derivatives thereof.", &
123 : usage="E_RANGE 10.0", &
124 64714 : default_r_val=-1.0_dp)
125 64714 : CALL section_add_keyword(section, keyword)
126 64714 : CALL keyword_release(keyword)
127 :
128 : CALL keyword_create( &
129 : keyword, __LOCATION__, &
130 : name="SCALE_S", &
131 : description="Scaling factor of the singlet energy component (opposite spin, OS) of the "// &
132 : "MP2, RI-MP2 and SOS-MP2 correlation energy. ", &
133 : usage="SCALE_S 1.0", &
134 64714 : default_r_val=1.0_dp)
135 64714 : CALL section_add_keyword(section, keyword)
136 64714 : CALL keyword_release(keyword)
137 :
138 : CALL keyword_create( &
139 : keyword, __LOCATION__, &
140 : name="SCALE_T", &
141 : description="Scaling factor of the triplet energy component (same spin, SS) of the MP2 "// &
142 : "and RI-MP2 correlation energy.", &
143 : usage="SCALE_T 1.0", &
144 64714 : default_r_val=1.0_dp)
145 64714 : CALL section_add_keyword(section, keyword)
146 64714 : CALL keyword_release(keyword)
147 :
148 : CALL keyword_create( &
149 : keyword, __LOCATION__, &
150 : name="GROUP_SIZE", &
151 : variants=(/"NUMBER_PROC"/), &
152 : description="Group size used in the computation of GPW and MME integrals and the MP2 correlation energy. "// &
153 : "The group size must be a divisor of the total number of MPI ranks. "// &
154 : "A smaller group size (for example the number of MPI ranks per node) "// &
155 : "accelerates the computation of integrals but a too large group size increases communication costs. "// &
156 : "A too small group size may lead to out of memory.", &
157 : usage="GROUP_SIZE 2", &
158 129428 : default_i_val=1)
159 64714 : CALL section_add_keyword(section, keyword)
160 64714 : CALL keyword_release(keyword)
161 :
162 64714 : NULLIFY (subsection)
163 64714 : CALL create_mp2_details_section(subsection)
164 64714 : CALL section_add_subsection(section, subsection)
165 64714 : CALL section_release(subsection)
166 :
167 64714 : CALL create_ri_mp2(subsection)
168 64714 : CALL section_add_subsection(section, subsection)
169 64714 : CALL section_release(subsection)
170 :
171 64714 : CALL create_ri_rpa(subsection)
172 64714 : CALL section_add_subsection(section, subsection)
173 64714 : CALL section_release(subsection)
174 :
175 64714 : CALL create_ri_laplace(subsection)
176 64714 : CALL section_add_subsection(section, subsection)
177 64714 : CALL section_release(subsection)
178 :
179 : ! here we generate an imag. time subsection to use with RPA or Laplace-SOS-MP2
180 64714 : CALL create_low_scaling(subsection)
181 64714 : CALL section_add_subsection(section, subsection)
182 64714 : CALL section_release(subsection)
183 :
184 64714 : CALL create_ri_section(subsection)
185 64714 : CALL section_add_subsection(section, subsection)
186 64714 : CALL section_release(subsection)
187 :
188 64714 : CALL create_integrals_section(subsection)
189 64714 : CALL section_add_subsection(section, subsection)
190 64714 : CALL section_release(subsection)
191 :
192 64714 : CALL create_canonical_gradients(subsection)
193 64714 : CALL section_add_subsection(section, subsection)
194 64714 : CALL section_release(subsection)
195 :
196 64714 : NULLIFY (print_key)
197 : CALL cp_print_key_section_create(print_key, __LOCATION__, "PRINT", &
198 : description="Controls the printing basic info about WFC methods", &
199 64714 : print_level=low_print_level, add_last=add_last_numeric, filename="__STD_OUT__")
200 64714 : CALL section_add_subsection(section, print_key)
201 64714 : CALL section_release(print_key)
202 :
203 64714 : END SUBROUTINE create_mp2_section
204 :
205 : ! **************************************************************************************************
206 : !> \brief ...
207 : !> \param section ...
208 : ! **************************************************************************************************
209 64714 : SUBROUTINE create_mp2_details_section(section)
210 : TYPE(section_type), POINTER :: section
211 :
212 : TYPE(keyword_type), POINTER :: keyword
213 :
214 64714 : CPASSERT(.NOT. ASSOCIATED(section))
215 : CALL section_create(section, __LOCATION__, name="MP2", &
216 : description="Parameters influencing MP2 (non-RI).", &
217 64714 : n_keywords=3, n_subsections=0, repeats=.FALSE.)
218 :
219 64714 : NULLIFY (keyword)
220 : CALL keyword_create(keyword, __LOCATION__, name="_SECTION_PARAMETERS_", &
221 : description="Activates MP2 calculations.", &
222 : usage="&MP2 .TRUE.", &
223 64714 : default_l_val=.FALSE., lone_keyword_l_val=.TRUE.)
224 64714 : CALL section_add_keyword(section, keyword)
225 64714 : CALL keyword_release(keyword)
226 :
227 : CALL keyword_create( &
228 : keyword, __LOCATION__, &
229 : name="METHOD", &
230 : citations=(/DelBen2012, DelBen2013/), &
231 : description="Method that is used to compute the MP2 energy.", &
232 : usage="METHOD MP2_GPW", &
233 : enum_c_vals=s2a("NONE", "DIRECT_CANONICAL", "MP2_GPW"), &
234 : enum_i_vals=(/mp2_method_none, mp2_method_direct, mp2_method_gpw/), &
235 : enum_desc=s2a("Skip MP2 calculation.", &
236 : "Use the direct mp2 canonical approach.", &
237 : "Use the GPW approach to MP2."), &
238 194142 : default_i_val=mp2_method_direct)
239 64714 : CALL section_add_keyword(section, keyword)
240 64714 : CALL keyword_release(keyword)
241 :
242 : CALL keyword_create( &
243 : keyword, __LOCATION__, &
244 : name="BIG_SEND", &
245 : description="Influencing the direct canonical MP2 method: Send big "// &
246 : "messages between processes (useful for >48 processors).", &
247 : usage="BIG_SEND", &
248 : default_l_val=.TRUE., &
249 64714 : lone_keyword_l_val=.TRUE.)
250 64714 : CALL section_add_keyword(section, keyword)
251 64714 : CALL keyword_release(keyword)
252 :
253 64714 : END SUBROUTINE create_mp2_details_section
254 :
255 : ! **************************************************************************************************
256 : !> \brief ...
257 : !> \param section ...
258 : ! **************************************************************************************************
259 64714 : SUBROUTINE create_ri_mp2(section)
260 : TYPE(section_type), POINTER :: section
261 :
262 : TYPE(keyword_type), POINTER :: keyword
263 :
264 64714 : CPASSERT(.NOT. ASSOCIATED(section))
265 : CALL section_create(section, __LOCATION__, name="RI_MP2", &
266 : description="Parameters influencing the RI-MP2 method. RI-MP2 supports gradients.", &
267 : n_keywords=3, n_subsections=1, repeats=.FALSE., &
268 129428 : citations=(/DelBen2013/))
269 :
270 64714 : NULLIFY (keyword)
271 : CALL keyword_create(keyword, __LOCATION__, name="_SECTION_PARAMETERS_", &
272 : description="Putting the &RI_MP2 section activates RI-MP2 calculation.", &
273 : usage="&RI_MP2 .TRUE.", &
274 64714 : default_l_val=.FALSE., lone_keyword_l_val=.TRUE.)
275 64714 : CALL section_add_keyword(section, keyword)
276 64714 : CALL keyword_release(keyword)
277 :
278 : CALL keyword_create(keyword, __LOCATION__, name="BLOCK_SIZE", &
279 : variants=(/"MESSAGE_SIZE"/), &
280 : description="Determines the blocking used for communication in RI-MP2. Larger BLOCK_SIZE "// &
281 : "reduces communication but requires more memory. The default (-1) is automatic.", &
282 : usage="BLOCK_SIZE 2", &
283 129428 : default_i_val=-1)
284 64714 : CALL section_add_keyword(section, keyword)
285 64714 : CALL keyword_release(keyword)
286 :
287 : CALL keyword_create(keyword, __LOCATION__, name="NUMBER_INTEGRATION_GROUPS", &
288 : description="Sets the number of integration groups of the communication scheme in RI-MP2. "// &
289 : "Integrals will be replicated such that each integration group has all integrals available. "// &
290 : "Must be a divisor of the number of subgroups (see GROUP_SIZE keyword in the WF_CORRELATION "// &
291 : "section. Smaller groups reduce the communication costs but increase the memory developments. "// &
292 : "If the provided value is non-positive or not a divisor of the number of subgroups, "// &
293 : "the number of integration groups is determined automatically (default).", &
294 : usage="NUMBER_INTEGRATION_GROUPS 2", &
295 64714 : default_i_val=-1)
296 64714 : CALL section_add_keyword(section, keyword)
297 64714 : CALL keyword_release(keyword)
298 :
299 : CALL keyword_create( &
300 : keyword, __LOCATION__, &
301 : name="PRINT_DGEMM_INFO", &
302 : description="Print details about all DGEMM calls.", &
303 : lone_keyword_l_val=.TRUE., &
304 64714 : default_l_val=.FALSE.)
305 64714 : CALL section_add_keyword(section, keyword)
306 64714 : CALL keyword_release(keyword)
307 :
308 64714 : END SUBROUTINE create_ri_mp2
309 :
310 : ! **************************************************************************************************
311 : !> \brief ...
312 : !> \param section ...
313 : ! **************************************************************************************************
314 64714 : SUBROUTINE create_opt_ri_basis(section)
315 : TYPE(section_type), POINTER :: section
316 :
317 : TYPE(keyword_type), POINTER :: keyword
318 :
319 64714 : CPASSERT(.NOT. ASSOCIATED(section))
320 : CALL section_create(section, __LOCATION__, name="OPT_RI_BASIS", &
321 : description="Parameters influencing the optimization of the RI MP2 basis. "// &
322 : "Only exponents of non-contracted auxiliary basis can be optimized. "// &
323 : "An initial RI auxiliary basis has to be specified.", &
324 : n_keywords=6, n_subsections=0, repeats=.FALSE., &
325 129428 : citations=(/DelBen2013/))
326 64714 : NULLIFY (keyword)
327 : CALL keyword_create(keyword, __LOCATION__, name="_SECTION_PARAMETERS_", &
328 : description="Putting the &OPT_RI_BASIS section activates optimization of RI basis.", &
329 : usage="&OPT_RI_BASIS .TRUE.", &
330 64714 : default_l_val=.FALSE., lone_keyword_l_val=.TRUE.)
331 64714 : CALL section_add_keyword(section, keyword)
332 64714 : CALL keyword_release(keyword)
333 :
334 : CALL keyword_create(keyword, __LOCATION__, name="DELTA_I_REL", &
335 : variants=(/"DI_REL"/), &
336 : description="Target accuracy in the relative deviation of the amplitudes calculated with "// &
337 : "and without RI approximation, (more details in Chem.Phys.Lett.294(1998)143).", &
338 : usage="DELTA_I_REL 1.0E-6_dp", &
339 129428 : default_r_val=1.0E-6_dp)
340 64714 : CALL section_add_keyword(section, keyword)
341 64714 : CALL keyword_release(keyword)
342 :
343 : CALL keyword_create(keyword, __LOCATION__, name="DELTA_RI", &
344 : variants=(/"DRI"/), &
345 : description="Target accuracy in the absolute difference between the RI-MP2 "// &
346 : "and the exact MP2 energy, DRI=ABS(E_MP2-E_RI-MP2).", &
347 : usage="DELTA_RI 1.0E-6_dp", &
348 129428 : default_r_val=5.0E-6_dp)
349 64714 : CALL section_add_keyword(section, keyword)
350 64714 : CALL keyword_release(keyword)
351 :
352 : CALL keyword_create(keyword, __LOCATION__, name="EPS_DERIV", &
353 : variants=(/"EPS_NUM_DERIV"/), &
354 : description="The derivatives of the MP2 energy with respect to the "// &
355 : "exponents of the basis are calculated numerically. "// &
356 : "The change in the exponent a_i employed for the numerical evaluation "// &
357 : "is defined as h_i=EPS_DERIV*a_i.", &
358 : usage="EPS_DERIV 1.0E-3_dp", &
359 129428 : default_r_val=1.0E-3_dp)
360 64714 : CALL section_add_keyword(section, keyword)
361 64714 : CALL keyword_release(keyword)
362 :
363 : CALL keyword_create(keyword, __LOCATION__, name="MAX_ITER", &
364 : variants=(/"MAX_NUM_ITER"/), &
365 : description="Specifies the maximum number of steps in the RI basis optimization.", &
366 : usage="MAX_ITER 100", &
367 129428 : default_i_val=50)
368 64714 : CALL section_add_keyword(section, keyword)
369 64714 : CALL keyword_release(keyword)
370 :
371 : CALL keyword_create(keyword, __LOCATION__, name="NUM_FUNC", &
372 : description="Specifies the number of function, for each angular momentum (s, p, d ...), "// &
373 : "employed in the automatically generated initial guess. "// &
374 : "This will be effective only if RI_AUX_BASIS_SET in the KIND section is not specified.", &
375 : usage="NUM_FUNC {number of s func.} {number of p func.} ...", &
376 64714 : n_var=-1, default_i_vals=(/-1/), type_of_var=integer_t)
377 64714 : CALL section_add_keyword(section, keyword)
378 64714 : CALL keyword_release(keyword)
379 :
380 : CALL keyword_create(keyword, __LOCATION__, name="BASIS_SIZE", &
381 : description="Specifies the size of the auxiliary basis set automatically "// &
382 : "generated as initial guess. This will be effective only if RI_AUX_BASIS_SET "// &
383 : "in the KIND section and NUM_FUNC are not specified.", &
384 : usage="BASIS_SIZE (MEDIUM|LARGE|VERY_LARGE)", &
385 : enum_c_vals=s2a("MEDIUM", "LARGE", "VERY_LARGE"), &
386 : enum_i_vals=(/0, 1, 2/), &
387 64714 : default_i_val=0)
388 64714 : CALL section_add_keyword(section, keyword)
389 64714 : CALL keyword_release(keyword)
390 :
391 64714 : END SUBROUTINE create_opt_ri_basis
392 :
393 : ! **************************************************************************************************
394 : !> \brief ...
395 : !> \param section ...
396 : ! **************************************************************************************************
397 64714 : SUBROUTINE create_ri_laplace(section)
398 : TYPE(section_type), POINTER :: section
399 :
400 : TYPE(keyword_type), POINTER :: keyword
401 :
402 64714 : CPASSERT(.NOT. ASSOCIATED(section))
403 : CALL section_create(section, __LOCATION__, name="RI_SOS_MP2", &
404 : description="Parameters influencing the RI-SOS-MP2-Laplace method", &
405 : n_keywords=3, n_subsections=1, repeats=.FALSE., &
406 129428 : citations=(/DelBen2013/))
407 :
408 64714 : NULLIFY (keyword)
409 : CALL keyword_create(keyword, __LOCATION__, name="_SECTION_PARAMETERS_", &
410 : description="Putting the &RI_SOS_MP2 section activates RI-SOS-MP2 calculation.", &
411 : usage="&RI_SOS_MP2 .TRUE.", &
412 64714 : default_l_val=.FALSE., lone_keyword_l_val=.TRUE.)
413 64714 : CALL section_add_keyword(section, keyword)
414 64714 : CALL keyword_release(keyword)
415 :
416 : CALL keyword_create( &
417 : keyword, __LOCATION__, name="QUADRATURE_POINTS", &
418 : variants=(/"LAPLACE_NUM_QUAD_POINTS"/), &
419 : description="Number of quadrature points for the numerical integration in the RI-SOS-MP2-Laplace method.", &
420 : usage="QUADRATURE_POINTS 6", &
421 129428 : default_i_val=5)
422 64714 : CALL section_add_keyword(section, keyword)
423 64714 : CALL keyword_release(keyword)
424 :
425 : CALL keyword_create( &
426 : keyword, __LOCATION__, name="NUM_INTEG_GROUPS", &
427 : description="Number of groups for the integration in the Laplace method. Each groups processes "// &
428 : "the same amount of quadrature points. It must be a divisor of the number of quadrature points and "// &
429 : "NUM_INTEG_GROUPS*GROUP_SIZE must be a divisor of the total number of processes. The default (-1) is automatic.", &
430 : usage="NUM_INTEG_GROUPS 2", &
431 64714 : default_i_val=-1)
432 64714 : CALL section_add_keyword(section, keyword)
433 64714 : CALL keyword_release(keyword)
434 :
435 64714 : END SUBROUTINE create_ri_laplace
436 :
437 : ! **************************************************************************************************
438 : !> \brief ...
439 : !> \param section ...
440 : ! **************************************************************************************************
441 64714 : SUBROUTINE create_canonical_gradients(section)
442 : TYPE(section_type), POINTER :: section
443 :
444 : TYPE(keyword_type), POINTER :: keyword
445 : TYPE(section_type), POINTER :: subsection
446 :
447 64714 : CPASSERT(.NOT. ASSOCIATED(section))
448 : CALL section_create(section, __LOCATION__, name="CANONICAL_GRADIENTS", &
449 : description="Parameters influencing gradient calculations of canonical RI methods. "// &
450 : "Ignored if the IM_TIME section is set.", &
451 : n_keywords=3, n_subsections=1, repeats=.FALSE., &
452 323570 : citations=(/DelBen2015b, Rybkin2016, Stein2022, Stein2024/))
453 :
454 64714 : NULLIFY (subsection, keyword)
455 64714 : CALL create_cphf(subsection)
456 64714 : CALL section_add_subsection(section, subsection)
457 64714 : CALL section_release(subsection)
458 :
459 : CALL keyword_create(keyword, __LOCATION__, name="EPS_CANONICAL", &
460 : description="Threshold under which a given ij or ab pair is considered to be degenerate and "// &
461 : "its contribution to the density matrix is calculated directly. "// &
462 : "Ignored in case of energy-only calculation.", &
463 : usage="EPS_CANONICAL 1.0E-8", type_of_var=real_t, &
464 64714 : default_r_val=1.0E-7_dp)
465 64714 : CALL section_add_keyword(section, keyword)
466 64714 : CALL keyword_release(keyword)
467 :
468 : CALL keyword_create( &
469 : keyword, __LOCATION__, &
470 : name="FREE_HFX_BUFFER", &
471 : description="Free the buffer containing the 4 center integrals used in the Hartree-Fock exchange calculation. "// &
472 : "Ignored for energy-only calculations. May fail.", &
473 : usage="FREE_HFX_BUFFER", &
474 : default_l_val=.FALSE., &
475 64714 : lone_keyword_l_val=.TRUE.)
476 64714 : CALL section_add_keyword(section, keyword)
477 64714 : CALL keyword_release(keyword)
478 :
479 : CALL keyword_create( &
480 : keyword, __LOCATION__, &
481 : name="DOT_PRODUCT_BLKSIZE", &
482 : description="Dot products for the calculation of the RPA/SOS-MP2 density matrices "// &
483 : "are calculated in batches of the size given by this keyword. Larger block sizes "// &
484 : "improve the performance but reduce the numerical accuracy. Recommended block sizes are multiples of the number of "// &
485 : "doubles per cache line (usually 8). Ignored with MP2 gradients. Set it to -1 to prevent blocking.", &
486 64714 : default_i_val=-1)
487 64714 : CALL section_add_keyword(section, keyword)
488 64714 : CALL keyword_release(keyword)
489 :
490 : CALL keyword_create( &
491 : keyword, __LOCATION__, &
492 : name="MAX_PARALLEL_COMM", &
493 : description="Sets the maximum number of parallel communication steps of the non-blocking communication scheme. "// &
494 : "The number of channels is determined from the available memory. If set to a value smaller than one, "// &
495 : "CP2K will use all memory for communication. A value of one enforces the blocking communication scheme "// &
496 : "increasing the communication costs.", &
497 64714 : default_i_val=2)
498 64714 : CALL section_add_keyword(section, keyword)
499 64714 : CALL keyword_release(keyword)
500 :
501 64714 : END SUBROUTINE create_canonical_gradients
502 :
503 : ! **************************************************************************************************
504 : !> \brief ...
505 : !> \param section ...
506 : ! **************************************************************************************************
507 64714 : SUBROUTINE create_ri_rpa(section)
508 : TYPE(section_type), POINTER :: section
509 :
510 : TYPE(keyword_type), POINTER :: keyword
511 : TYPE(section_type), POINTER :: subsection
512 :
513 64714 : CPASSERT(.NOT. ASSOCIATED(section))
514 : CALL section_create(section, __LOCATION__, name="RI_RPA", &
515 : description="Parameters influencing RI-RPA and GW.", &
516 : n_keywords=8, n_subsections=4, repeats=.FALSE., &
517 194142 : citations=(/DelBen2013, DelBen2015/))
518 :
519 64714 : NULLIFY (keyword, subsection)
520 : CALL keyword_create(keyword, __LOCATION__, name="_SECTION_PARAMETERS_", &
521 : description="Putting the &RI_RPA section activates RI-RPA calculation.", &
522 : usage="&RI_RPA .TRUE.", &
523 64714 : default_l_val=.FALSE., lone_keyword_l_val=.TRUE.)
524 64714 : CALL section_add_keyword(section, keyword)
525 64714 : CALL keyword_release(keyword)
526 :
527 : CALL keyword_create( &
528 : keyword, __LOCATION__, &
529 : name="SIGMA_FUNCTIONAL", &
530 : description="Determine parametrization for sigma-functional", &
531 : usage="SIGMA_FUNCTIONAL PBE_S2", &
532 : enum_c_vals=s2a("NONE", "PBE0_S1", "PBE0_S2", "PBE_S1", "PBE_S2"), &
533 : enum_i_vals=(/sigma_none, sigma_PBE0_S1, sigma_PBE0_S2, sigma_PBE_S1, sigma_PBE_S2/), &
534 : enum_desc=s2a("No sigma functional calculation", &
535 : "use parameters based on PBE0 with S1 set.", &
536 : "use parameters based on PBE0 with S2 set.", &
537 : "use parameters based on PBE with S1 set.", &
538 : "use parameters based on PBE with S2 set." &
539 : ), &
540 64714 : default_i_val=sigma_none)
541 64714 : CALL section_add_keyword(section, keyword)
542 64714 : CALL keyword_release(keyword)
543 :
544 : CALL keyword_create(keyword, __LOCATION__, name="QUADRATURE_POINTS", &
545 : variants=(/"RPA_NUM_QUAD_POINTS"/), &
546 : description="Number of quadrature points for the numerical integration in the RI-RPA method.", &
547 : usage="QUADRATURE_POINTS 60", &
548 129428 : default_i_val=40)
549 64714 : CALL section_add_keyword(section, keyword)
550 64714 : CALL keyword_release(keyword)
551 :
552 : CALL keyword_create(keyword, __LOCATION__, name="NUM_INTEG_GROUPS", &
553 : description="Number of groups for the integration in the Laplace method. Each groups processes "// &
554 : "the same amount of quadrature points. It must be a divisor of the number of quadrature points and "// &
555 : "NUM_INTEG_GROUPS*GROUP_SIZE must be a divisor of the total number of processes. "// &
556 : "The default (-1) is automatic.", &
557 : usage="NUM_INTEG_GROUPS 2", &
558 64714 : default_i_val=-1)
559 64714 : CALL section_add_keyword(section, keyword)
560 64714 : CALL keyword_release(keyword)
561 :
562 : CALL keyword_create(keyword, __LOCATION__, &
563 : name="MM_STYLE", &
564 : description="Matrix multiplication style for the Q matrix.", &
565 : usage="MM_STYLE GEMM", &
566 : enum_c_vals=s2a("GEMM", "SYRK"), &
567 : enum_i_vals=(/wfc_mm_style_gemm, wfc_mm_style_syrk/), &
568 : enum_desc=s2a("Use pdgemm: more flops, maybe faster.", &
569 : "Use pdysrk: fewer flops, maybe slower."), &
570 64714 : default_i_val=wfc_mm_style_gemm)
571 64714 : CALL section_add_keyword(section, keyword)
572 64714 : CALL keyword_release(keyword)
573 :
574 : CALL keyword_create( &
575 : keyword, __LOCATION__, &
576 : name="MINIMAX_QUADRATURE", &
577 : variants=(/"MINIMAX"/), &
578 : description="Use the Minimax quadrature scheme for performing the numerical integration. "// &
579 : "Maximum number of quadrature point limited to 20.", &
580 : usage="MINIMAX_QUADRATURE", &
581 : default_l_val=.FALSE., &
582 129428 : lone_keyword_l_val=.TRUE.)
583 64714 : CALL section_add_keyword(section, keyword)
584 64714 : CALL keyword_release(keyword)
585 :
586 : CALL keyword_create( &
587 : keyword, __LOCATION__, &
588 : name="RSE", &
589 : variants=(/"SE"/), &
590 : description="Decide whether to add singles correction.", &
591 : usage="RSE", &
592 : default_l_val=.FALSE., &
593 129428 : lone_keyword_l_val=.TRUE.)
594 64714 : CALL section_add_keyword(section, keyword)
595 64714 : CALL keyword_release(keyword)
596 :
597 : CALL keyword_create( &
598 : keyword, __LOCATION__, &
599 : name="ADMM", &
600 : description="Decide whether to perform ADMM in the exact exchange calc. for RPA and/or GW. "// &
601 : "The ADMM XC correction is governed by the AUXILIARY_DENSITY_MATRIX_METHOD section in &DFT. "// &
602 : "In most cases, the Hartree-Fock exchange is not too expensive and there is no need for ADMM, "// &
603 : "ADMM can however provide significant speedup and memory savings in case of diffuse basis sets. "// &
604 : "If it is a GW bandgap calculations, RI_SIGMA_X can also be used. ", &
605 : usage="ADMM", &
606 : default_l_val=.FALSE., &
607 64714 : lone_keyword_l_val=.TRUE.)
608 64714 : CALL section_add_keyword(section, keyword)
609 64714 : CALL keyword_release(keyword)
610 :
611 : CALL keyword_create( &
612 : keyword, __LOCATION__, &
613 : name="SCALE_RPA", &
614 : description="Scales RPA energy contributions (RPA, exchange correction).", &
615 : usage="SCALE_RPA 1.0", &
616 64714 : default_r_val=1.0_dp)
617 64714 : CALL section_add_keyword(section, keyword)
618 64714 : CALL keyword_release(keyword)
619 :
620 : CALL keyword_create( &
621 : keyword, __LOCATION__, &
622 : name="PRINT_DGEMM_INFO", &
623 : description="Print details about all DGEMM calls.", &
624 : lone_keyword_l_val=.TRUE., &
625 64714 : default_l_val=.FALSE.)
626 64714 : CALL section_add_keyword(section, keyword)
627 64714 : CALL keyword_release(keyword)
628 :
629 : ! here we generate a hfx subsection to use in the case EXX has to be computed after RPA
630 64714 : CALL create_hfx_section(subsection)
631 64714 : CALL section_add_subsection(section, subsection)
632 64714 : CALL section_release(subsection)
633 :
634 : ! here we generate a G0W0 subsection to use if G0W0 is desired
635 64714 : CALL create_ri_g0w0(subsection)
636 64714 : CALL section_add_subsection(section, subsection)
637 64714 : CALL section_release(subsection)
638 :
639 : ! here we the RPA exchange section
640 64714 : CALL create_rpa_exchange(subsection)
641 64714 : CALL section_add_subsection(section, subsection)
642 64714 : CALL section_release(subsection)
643 :
644 64714 : END SUBROUTINE create_ri_rpa
645 :
646 : ! **************************************************************************************************
647 : !> \brief ...
648 : !> \param section ...
649 : ! **************************************************************************************************
650 64714 : SUBROUTINE create_rpa_exchange(section)
651 : TYPE(section_type), POINTER :: section
652 :
653 : TYPE(keyword_type), POINTER :: keyword
654 :
655 64714 : CPASSERT(.NOT. ASSOCIATED(section))
656 : CALL section_create(section, __LOCATION__, name="EXCHANGE_CORRECTION", &
657 : description="Parameters influencing exchange corrections to RPA. No gradients available.", &
658 64714 : n_keywords=3, n_subsections=1, repeats=.FALSE.)
659 :
660 64714 : NULLIFY (keyword)
661 : CALL keyword_create(keyword, __LOCATION__, name="_SECTION_PARAMETERS_", &
662 : description="Choose the kind of exchange correction.", &
663 : usage="&EXCHANGE_CORRECTION AXK", &
664 : enum_c_vals=s2a("NONE", "AXK", "SOSEX"), &
665 : enum_i_vals=(/rpa_exchange_none, rpa_exchange_axk, rpa_exchange_sosex/), &
666 : enum_desc=s2a("Apply no exchange correction.", &
667 : "Apply Approximate eXchange Kernel (AXK) correction.", &
668 : "Apply Second Order Screened eXchange (SOSEX) correction."), &
669 64714 : default_i_val=rpa_exchange_none)
670 64714 : CALL section_add_keyword(section, keyword)
671 64714 : CALL keyword_release(keyword)
672 :
673 : CALL keyword_create( &
674 : keyword, __LOCATION__, &
675 : name="BLOCK_SIZE", &
676 : description="Choose the block size of the contraction step. Larger block sizes improve performance but "// &
677 : "require more memory (quadratically!, number of stored elements: $o^2\cdot N_B^2$). "// &
678 : "Nonpositive numbers turn off blocking.", &
679 : usage="BLOCK_SIZE 1", &
680 64714 : default_i_val=1)
681 64714 : CALL section_add_keyword(section, keyword)
682 64714 : CALL keyword_release(keyword)
683 :
684 : CALL keyword_create( &
685 : keyword, __LOCATION__, &
686 : name="USE_HFX_IMPLEMENTATION", &
687 : description="Use a HF-based implementation with RI_RPA%HF section. Recommended for large systems.", &
688 : usage="USE_HFX_IMPLEMENTATION T", &
689 64714 : default_l_val=.FALSE., lone_keyword_l_val=.TRUE.)
690 64714 : CALL section_add_keyword(section, keyword)
691 64714 : CALL keyword_release(keyword)
692 :
693 64714 : END SUBROUTINE create_rpa_exchange
694 :
695 : ! **************************************************************************************************
696 : !> \brief ...
697 : !> \param section ...
698 : ! **************************************************************************************************
699 64714 : SUBROUTINE create_ri_g0w0(section)
700 : TYPE(section_type), POINTER :: section
701 :
702 : TYPE(keyword_type), POINTER :: keyword
703 : TYPE(section_type), POINTER :: subsection
704 :
705 64714 : CPASSERT(.NOT. ASSOCIATED(section))
706 : CALL section_create(section, __LOCATION__, name="GW", &
707 : description="Parameters influencing GW calculations on molecules, "// &
708 : "see also 'Electronic band structure from GW', "// &
709 : "https://manual.cp2k.org/trunk/methods/properties/bandstructure_gw.html.", &
710 64714 : n_keywords=24, n_subsections=1, repeats=.FALSE.)
711 :
712 64714 : NULLIFY (keyword, subsection)
713 :
714 : CALL keyword_create(keyword, __LOCATION__, name="_SECTION_PARAMETERS_", &
715 : description="Activates GW calculations.", &
716 : usage="&GW .TRUE.", &
717 64714 : default_l_val=.FALSE., lone_keyword_l_val=.TRUE.)
718 64714 : CALL section_add_keyword(section, keyword)
719 64714 : CALL keyword_release(keyword)
720 :
721 : CALL keyword_create(keyword, __LOCATION__, name="SELF_CONSISTENCY", &
722 : description="Decide the level of self-consistency of eigenvalues "// &
723 : "(= quasiparticle energies = single-electron energies) in GW. "// &
724 : "Updates of Kohn-Sham orbitals (for example qsGW) are not implemented. "// &
725 : "For details which type of eigenvalue self-consistency might be good, "// &
726 : "please consult Golze, Dvorak, Rinke, Front. Chem. 2019.", &
727 : usage="SELF_CONSISTENCY evGW0", &
728 : enum_c_vals=s2a("G0W0", "evGW0", "evGW"), &
729 : enum_i_vals=(/G0W0, evGW0, evGW/), &
730 : enum_desc=s2a("Use DFT eigenvalues; not update.", &
731 : "Update DFT eigenvalues in G, not in W.", &
732 : "Update DFT eigenvalues in G and W."), &
733 64714 : default_i_val=G0W0)
734 64714 : CALL section_add_keyword(section, keyword)
735 64714 : CALL keyword_release(keyword)
736 :
737 : CALL keyword_create(keyword, __LOCATION__, name="CORR_MOS_OCC", &
738 : variants=(/"CORR_OCC"/), &
739 : description="Number of occupied MOs whose energies are corrected in GW. "// &
740 : "Counting beginning from HOMO, e.g. 3 corrected occ. MOs correspond "// &
741 : "to correction of HOMO, HOMO-1 and HOMO-2. Numerical effort and "// &
742 : "storage of RI-G0W0 increase linearly with this number. In case you "// &
743 : "want to correct all occ. MOs, insert either a negative number or "// &
744 : "a number larger than the number of occ. MOs. Invoking CORR_MOS_OCC -2 "// &
745 : "together with a BSE cutoff, sets a sufficiently large CORR_MOS_OCC "// &
746 : "for the given BSE cutoff deduced from DFT eigenvalues.", &
747 : usage="CORR_OCC 3", &
748 129428 : default_i_val=10)
749 64714 : CALL section_add_keyword(section, keyword)
750 64714 : CALL keyword_release(keyword)
751 :
752 : CALL keyword_create(keyword, __LOCATION__, name="CORR_MOS_VIRT", &
753 : variants=(/"CORR_VIRT"/), &
754 : description="Number of virtual MOs whose energies are corrected by GW. "// &
755 : "Counting beginning from LUMO, e.g. 3 corrected occ. MOs correspond "// &
756 : "to correction of LUMO, LUMO+1 and LUMO+2. Numerical effort and "// &
757 : "storage of RI-G0W0 increase linearly with this number. In case you "// &
758 : "want to correct all virt. MOs, insert either a negative number or "// &
759 : "a number larger than the number of virt. MOs. Invoking CORR_MOS_VIRT -2 "// &
760 : "together with a BSE cutoff, sets a sufficiently large CORR_MOS_VIRT "// &
761 : "for the given BSE cutoff deduced from DFT eigenvalues.", &
762 : usage="CORR_VIRT 3", &
763 129428 : default_i_val=10)
764 64714 : CALL section_add_keyword(section, keyword)
765 64714 : CALL keyword_release(keyword)
766 :
767 : CALL keyword_create(keyword, __LOCATION__, name="NUMB_POLES", &
768 : description="Number of poles for the fitting. Usually, two poles are sufficient. ", &
769 : usage="NUMB_POLES 2", &
770 64714 : default_i_val=2)
771 64714 : CALL section_add_keyword(section, keyword)
772 64714 : CALL keyword_release(keyword)
773 :
774 : CALL keyword_create(keyword, __LOCATION__, name="OMEGA_MAX_FIT", &
775 : description="Determines fitting range for the self-energy on the imaginary axis: "// &
776 : "[0, OMEGA_MAX_FIT] for virt orbitals, [-OMEGA_MAX_FIT,0] for occ orbitals. "// &
777 : "Unit: Hartree. Default: 0.734996 H = 20 eV. ", &
778 : usage="OMEGA_MAX_FIT 0.5", &
779 64714 : default_r_val=0.734996_dp)
780 64714 : CALL section_add_keyword(section, keyword)
781 64714 : CALL keyword_release(keyword)
782 :
783 : CALL keyword_create(keyword, __LOCATION__, name="CROSSING_SEARCH", &
784 : description="Determines, how the self_energy is evaluated on the real axis.", &
785 : usage="CROSSING_SEARCH Z_SHOT", &
786 : enum_c_vals=s2a("Z_SHOT", "NEWTON", "BISECTION"), &
787 : enum_i_vals=(/ri_rpa_g0w0_crossing_z_shot, &
788 : ri_rpa_g0w0_crossing_newton, ri_rpa_g0w0_crossing_bisection/), &
789 : enum_desc=s2a("Calculate the derivative of Sigma and out of it Z. Then extrapolate using Z.", &
790 : "Make a Newton-Raphson fix point iteration.", &
791 : "Make a bisection fix point iteration."), &
792 64714 : default_i_val=ri_rpa_g0w0_crossing_newton)
793 64714 : CALL section_add_keyword(section, keyword)
794 64714 : CALL keyword_release(keyword)
795 :
796 : CALL keyword_create(keyword, __LOCATION__, name="FERMI_LEVEL_OFFSET", &
797 : description="Fermi level for occ. orbitals: e_HOMO + FERMI_LEVEL_OFFSET; "// &
798 : "Fermi level for virt. orbitals: e_LUMO - FERMI_LEVEL_OFFSET. "// &
799 : "In case e_homo + FERMI_LEVEL_OFFSET < e_lumo - FERMI_LEVEL_OFFSET, "// &
800 : "we set Fermi level = (e_HOMO+e_LUMO)/2. For cubic-scaling GW, the Fermi level "// &
801 : "is always equal to (e_HOMO+e_LUMO)/2 regardless of FERMI_LEVEL_OFFSET.", &
802 : usage="FERMI_LEVEL_OFFSET 1.0E-2", &
803 64714 : default_r_val=2.0E-2_dp)
804 64714 : CALL section_add_keyword(section, keyword)
805 64714 : CALL keyword_release(keyword)
806 :
807 : CALL keyword_create(keyword, __LOCATION__, name="HEDIN_SHIFT", &
808 : description="If true, use Hedin's shift in G0W0, evGW and evGW0 "// &
809 : "(aka scGW0). Details see in Li et al. JCTC 18, 7570 "// &
810 : "(2022), Figure 1. G0W0 with Hedin's shift should give "// &
811 : "similar GW eigenvalues as evGW0; at a lower "// &
812 : "computational cost.", &
813 : usage="HEDIN_SHIFT", &
814 : default_l_val=.FALSE., &
815 64714 : lone_keyword_l_val=.TRUE.)
816 64714 : CALL section_add_keyword(section, keyword)
817 64714 : CALL keyword_release(keyword)
818 :
819 : CALL keyword_create(keyword, __LOCATION__, name="EV_GW_ITER", &
820 : description="Maximum number of iterations for eigenvalue "// &
821 : "self-consistency cycle. The computational effort of GW scales "// &
822 : "linearly with this number. In case of choosing "// &
823 : "GW_SELF_CONSISTENCY EVGW, the code sets EV_GW_ITER 10.", &
824 : usage="EV_GW_ITER 3", &
825 64714 : default_i_val=1)
826 64714 : CALL section_add_keyword(section, keyword)
827 64714 : CALL keyword_release(keyword)
828 :
829 : CALL keyword_create(keyword, __LOCATION__, name="SC_GW0_ITER", &
830 : description="Maximum number of iterations for GW0 "// &
831 : "self-consistency cycle. The computational effort "// &
832 : "of GW is not much affected by the number of scGW0 cycles. "// &
833 : "In case of choosing "// &
834 : "GW_SELF_CONSISTENCY EVGW0, the code sets SC_GW0_ITER 10.", &
835 : usage="SC_GW0_ITER 3", &
836 64714 : default_i_val=1)
837 64714 : CALL section_add_keyword(section, keyword)
838 64714 : CALL keyword_release(keyword)
839 :
840 : CALL keyword_create(keyword, __LOCATION__, name="EPS_ITER", &
841 : description="Target accuracy for the eigenvalue self-consistency. "// &
842 : "If the G0W0 HOMO-LUMO gap differs by less than the "// &
843 : "target accuracy during the iteration, the eigenvalue "// &
844 : "self-consistency cycle stops. Unit: Hartree.", &
845 : usage="EPS_ITER 0.00005", &
846 : default_r_val=cp_unit_to_cp2k(value=0.00136_dp, unit_str="eV"), &
847 64714 : unit_str="eV")
848 :
849 64714 : CALL section_add_keyword(section, keyword)
850 64714 : CALL keyword_release(keyword)
851 :
852 : CALL keyword_create(keyword, __LOCATION__, name="PRINT_EXX", &
853 : description="Print exchange self-energy minus exchange correlation potential for Gamma-only "// &
854 : "calculation (PRINT). For a GW calculation with k-points we use this output as "// &
855 : "exchange self-energy (READ). This is a temporary solution because the hybrid MPI/OMP "// &
856 : "parallelization in the HFX by Manuel Guidon conflicts with the parallelization in "// &
857 : "low-scaling GW k-points which is most efficient with maximum number of MPI tasks and "// &
858 : "minimum number of OMP threads. For HFX by M. Guidon, the density matrix is "// &
859 : "fully replicated on every MPI rank which necessitates a high number of OMP threads per MPI "// &
860 : "rank for large systems to prevent out of memory. "// &
861 : "Such a high number of OMP threads would slow down the GW calculation "// &
862 : "severely. Therefore, it was decided to temporarily divide the GW k-point calculation in a "// &
863 : "Gamma-only HF calculation with high number of OMP threads to prevent out of memory and "// &
864 : "a GW k-point calculation with 1 OMP thread per MPI rank reading the previousHF output.", &
865 : usage="PRINT_EXX TRUE", &
866 : enum_c_vals=s2a("TRUE", "FALSE", "READ", "SKIP_FOR_REGTEST"), &
867 : enum_i_vals=(/gw_print_exx, gw_no_print_exx, gw_read_exx, gw_skip_for_regtest/), &
868 : enum_desc=s2a("Please, put TRUE for Gamma only calculation to get the exchange self-energy. "// &
869 : "If 'SIGMA_X' and the corresponding values for the exchange-energy are written, "// &
870 : "the writing has been successful", &
871 : "FALSE is needed if you want to do nothing here.", &
872 : "Please, put READ for the k-point GW calculation to read the exact exchange. "// &
873 : "You have to provide an output file including the exact exchange. This file "// &
874 : "has to be named 'exx.dat'.", &
875 : "SKIP_FOR_REGTEST is only used for the GW k-point regtest where no exchange "// &
876 : "self-energy is computed."), &
877 64714 : default_i_val=gw_no_print_exx)
878 64714 : CALL section_add_keyword(section, keyword)
879 64714 : CALL keyword_release(keyword)
880 :
881 : CALL keyword_create(keyword, __LOCATION__, name="PRINT_SELF_ENERGY", &
882 : description="If true, print the self-energy for all levels for real energy "// &
883 : "together with the straight line to see the quasiparticle energy as intersection. "// &
884 : "In addition, prints the self-energy for imaginary frequencies together with the Pade fit.", &
885 : usage="PRINT_SELF_ENERGY", &
886 : default_l_val=.FALSE., &
887 64714 : lone_keyword_l_val=.TRUE.)
888 64714 : CALL section_add_keyword(section, keyword)
889 64714 : CALL keyword_release(keyword)
890 :
891 : CALL keyword_create(keyword, __LOCATION__, name="RI_SIGMA_X", &
892 : description="If true, the exchange self-energy is calculated approximatively with RI. "// &
893 : "If false, the Hartree-Fock implementation in CP2K is used.", &
894 : usage="RI_SIGMA_X", &
895 : default_l_val=.TRUE., &
896 64714 : lone_keyword_l_val=.TRUE.)
897 64714 : CALL section_add_keyword(section, keyword)
898 64714 : CALL keyword_release(keyword)
899 :
900 : CALL keyword_create(keyword, __LOCATION__, name="IC_CORR_LIST", &
901 : description="List of image charge correction from a previous calculation to be applied in G0W0 "// &
902 : "or evGW. Keyword is active, if the first entry is positive (since IC corrections are positive "// &
903 : "occupied MOs. The start corresponds to the first corrected GW level.", &
904 : usage="IC_CORR_LIST <REAL> ... <REAL>", &
905 : default_r_vals=(/-1.0_dp/), &
906 64714 : type_of_var=real_t, n_var=-1, unit_str="eV")
907 64714 : CALL section_add_keyword(section, keyword)
908 64714 : CALL keyword_release(keyword)
909 :
910 : CALL keyword_create(keyword, __LOCATION__, name="IC_CORR_LIST_BETA", &
911 : description="IC_CORR_LIST for beta spins in case of open shell calculation.", &
912 : usage="IC_CORR_LIST_BETA <REAL> ... <REAL>", &
913 : default_r_vals=(/-1.0_dp/), &
914 64714 : type_of_var=real_t, n_var=-1, unit_str="eV")
915 64714 : CALL section_add_keyword(section, keyword)
916 64714 : CALL keyword_release(keyword)
917 :
918 : CALL keyword_create(keyword, __LOCATION__, name="PERIODIC_CORRECTION", &
919 : description="If true, the periodic correction scheme is used employing k-points. "// &
920 : "Method is not recommended to use, use instead PERIODIC_LOW_SCALING which much "// &
921 : "more accurate than the periodic correction.", &
922 : usage="PERIODIC_CORRECTION", &
923 : default_l_val=.FALSE., &
924 64714 : lone_keyword_l_val=.TRUE.)
925 64714 : CALL section_add_keyword(section, keyword)
926 64714 : CALL keyword_release(keyword)
927 :
928 : CALL keyword_create(keyword, __LOCATION__, name="IMAGE_CHARGE_MODEL", &
929 : variants=(/"IC"/), &
930 : description="If true, an image charge model is applied to mimic the renormalization of "// &
931 : "electronic levels of a molecule at a metallic surface. For this calculation, the molecule "// &
932 : "has to be reflected on the desired xy image plane. The coordinates of the reflected molecule "// &
933 : "have to be added to the coord file as ghost atoms. For the ghost atoms, identical basis sets "// &
934 : "the normal atoms have to be used.", &
935 : usage="IC TRUE", &
936 : default_l_val=.FALSE., &
937 129428 : lone_keyword_l_val=.TRUE.)
938 64714 : CALL section_add_keyword(section, keyword)
939 64714 : CALL keyword_release(keyword)
940 :
941 : CALL keyword_create(keyword, __LOCATION__, name="ANALYTIC_CONTINUATION", &
942 : description="Defines which type of analytic continuation for the self energy is used", &
943 : usage="ANALYTIC_CONTINUATION", &
944 : enum_c_vals=s2a("TWO_POLE", "PADE"), &
945 : enum_i_vals=(/gw_two_pole_model, gw_pade_approx/), &
946 : enum_desc=s2a("Use 'two-pole' model.", &
947 : "Use Pade approximation."), &
948 64714 : default_i_val=gw_pade_approx)
949 64714 : CALL section_add_keyword(section, keyword)
950 64714 : CALL keyword_release(keyword)
951 :
952 : CALL keyword_create(keyword, __LOCATION__, name="NPARAM_PADE", &
953 : description="Number of parameters for the Pade approximation "// &
954 : "when using the latter for the analytic continuation of the "// &
955 : "self energy. 16 parameters (corresponding to 8 poles) are "// &
956 : "are recommended.", &
957 : usage="NPARAM_PADE 16", &
958 64714 : default_i_val=16)
959 64714 : CALL section_add_keyword(section, keyword)
960 64714 : CALL keyword_release(keyword)
961 :
962 : CALL keyword_create(keyword, __LOCATION__, name="GAMMA_ONLY_SIGMA", &
963 : variants=(/"GAMMA"/), &
964 : description="If true, the correlation self-energy is only computed at the Gamma point. "// &
965 : "The Gamma point itself is obtained by averaging over all kpoints of the DFT mesh.", &
966 : usage="GAMMA TRUE", &
967 : default_l_val=.FALSE., &
968 129428 : lone_keyword_l_val=.TRUE.)
969 64714 : CALL section_add_keyword(section, keyword)
970 64714 : CALL keyword_release(keyword)
971 :
972 : CALL keyword_create(keyword, __LOCATION__, name="UPDATE_XC_ENERGY", &
973 : description="If true, the Hartree-Fock and RPA total energy are printed and the total energy "// &
974 : "is corrected using exact exchange and the RPA correlation energy.", &
975 : usage="UPDATE_XC_ENERGY", &
976 : default_l_val=.FALSE., &
977 64714 : lone_keyword_l_val=.TRUE.)
978 64714 : CALL section_add_keyword(section, keyword)
979 64714 : CALL keyword_release(keyword)
980 :
981 : CALL keyword_create(keyword, __LOCATION__, name="KPOINTS_SELF_ENERGY", &
982 : description="Specify number of k-points for the k-point grid of the self-energy. Internally, a "// &
983 : "Monkhorst-Pack grid is used. A dense k-point grid may be necessary to compute an accurate density "// &
984 : "of state from GW. Large self-energy k-meshes do not cost much more computation time.", &
985 : usage="KPOINTS_SELF_ENERGY nx ny nz", repeats=.TRUE., &
986 64714 : n_var=3, type_of_var=integer_t, default_i_vals=(/0, 0, 0/))
987 64714 : CALL section_add_keyword(section, keyword)
988 64714 : CALL keyword_release(keyword)
989 :
990 : CALL keyword_create(keyword, __LOCATION__, name="REGULARIZATION_MINIMAX", &
991 : description="Tikhonov regularization for computing weights of the Fourier transform "// &
992 : "from imaginary time to imaginary frequency and vice versa. Needed for large minimax "// &
993 : "grids with 20 or more points and a small range.", &
994 : usage="REGULARIZATION_MINIMAX 1.0E-6", &
995 64714 : default_r_val=0.0_dp)
996 64714 : CALL section_add_keyword(section, keyword)
997 64714 : CALL keyword_release(keyword)
998 :
999 : CALL keyword_create(keyword, __LOCATION__, name="SOC", &
1000 : description="Calculate the spin-orbit splitting of the eigenvalues/band structure "// &
1001 : "using the spin-orbit part of the GTH pseudos parametrized in Hartwigsen, Goedecker, "// &
1002 : "Hutter, Phys. Rev. B 58, 3641 (1998), Eq. 19, "// &
1003 : "parameters in Table I.", &
1004 : usage="SOC", &
1005 : enum_c_vals=s2a("NONE", "LDA", "PBE"), &
1006 : enum_i_vals=(/soc_none, soc_lda, soc_pbe/), &
1007 : enum_desc=s2a("No SOC.", &
1008 : "Use parameters from LDA (PADE) pseudopotential.", &
1009 : "Use parameters from PBE pseudopotential."), &
1010 64714 : default_i_val=soc_none)
1011 64714 : CALL section_add_keyword(section, keyword)
1012 64714 : CALL keyword_release(keyword)
1013 :
1014 : CALL keyword_create(keyword, __LOCATION__, name="SOC_ENERGY_WINDOW", &
1015 : description="For perturbative SOC calculation, only "// &
1016 : "take frontier levels in an energy window "// &
1017 : "[E_HOMO - SOC_ENERGY_WINDOW/2 , E_LUMO + SOC_ENERGY_WINDOW/2 "// &
1018 : "into account for the diagonalization of H^GW,SOC.", &
1019 : usage="SOC_ENERGY_WINDOW 20.0_eV", &
1020 : default_r_val=cp_unit_to_cp2k(value=50.0_dp, unit_str="eV"), &
1021 64714 : unit_str="eV")
1022 64714 : CALL section_add_keyword(section, keyword)
1023 64714 : CALL keyword_release(keyword)
1024 :
1025 : ! here we generate a subsection for the periodic GW correction
1026 64714 : CALL create_periodic_gw_correction_section(subsection)
1027 64714 : CALL section_add_subsection(section, subsection)
1028 64714 : CALL section_release(subsection)
1029 :
1030 : ! here we generate a subsection for Bethe-Salpeter
1031 64714 : CALL create_bse_section(subsection)
1032 64714 : CALL section_add_subsection(section, subsection)
1033 64714 : CALL section_release(subsection)
1034 :
1035 : ! here we generate a subsection for image charge calculations
1036 64714 : CALL create_ic_section(subsection)
1037 64714 : CALL section_add_subsection(section, subsection)
1038 64714 : CALL section_release(subsection)
1039 :
1040 : ! here we generate a subsection for calculating the GW band structures
1041 64714 : CALL create_kpoint_set_section(subsection)
1042 64714 : CALL section_add_subsection(section, subsection)
1043 64714 : CALL section_release(subsection)
1044 :
1045 : ! here we generate a subsection for additional printing
1046 64714 : CALL create_print_section(subsection)
1047 64714 : CALL section_add_subsection(section, subsection)
1048 64714 : CALL section_release(subsection)
1049 :
1050 64714 : END SUBROUTINE create_ri_g0w0
1051 :
1052 : ! **************************************************************************************************
1053 : !> \brief ...
1054 : !> \param section ...
1055 : ! **************************************************************************************************
1056 64714 : SUBROUTINE create_print_section(section)
1057 : TYPE(section_type), POINTER :: section
1058 :
1059 : TYPE(keyword_type), POINTER :: keyword
1060 : TYPE(section_type), POINTER :: gw_dos_section, print_key
1061 :
1062 64714 : CPASSERT(.NOT. ASSOCIATED(section))
1063 64714 : NULLIFY (print_key, keyword)
1064 64714 : NULLIFY (gw_dos_section, keyword)
1065 : CALL section_create(section, __LOCATION__, name="PRINT", &
1066 : description="Section of possible print options specific for the GW code.", &
1067 64714 : n_keywords=0, n_subsections=2, repeats=.FALSE.)
1068 :
1069 : CALL cp_print_key_section_create(print_key, __LOCATION__, "LOCAL_BANDGAP", &
1070 : description="Prints a local bandgap E_gap(r), derived from the local density of "// &
1071 : "states rho(r,E). Details and formulae in the SI of the periodic GW paper (2023).", &
1072 : print_level=high_print_level, add_last=add_last_numeric, &
1073 : filename="LOCAL_BANDGAP", &
1074 64714 : common_iter_levels=3)
1075 :
1076 : CALL keyword_create(keyword, __LOCATION__, name="ENERGY_WINDOW", &
1077 : description="Energy window in the LDOS for searching the gap.", &
1078 : usage="ENERGY_WINDOW 6.0", &
1079 : default_r_val=cp_unit_to_cp2k(value=6.0_dp, unit_str="eV"), &
1080 64714 : unit_str="eV")
1081 64714 : CALL section_add_keyword(print_key, keyword)
1082 64714 : CALL keyword_release(keyword)
1083 :
1084 : CALL keyword_create(keyword, __LOCATION__, name="ENERGY_SPACING", &
1085 : description="Energy spacing of the LDOS for searching the gap.", &
1086 : usage="ENERGY_SPACING 0.03", &
1087 : default_r_val=cp_unit_to_cp2k(value=0.03_dp, unit_str="eV"), &
1088 64714 : unit_str="eV")
1089 64714 : CALL section_add_keyword(print_key, keyword)
1090 64714 : CALL keyword_release(keyword)
1091 :
1092 : CALL keyword_create(keyword, __LOCATION__, name="LDOS_THRESHOLD_GAP", &
1093 : description="Relative LDOS threshold that determines the local bandgap.", &
1094 : usage="LDOS_THRESHOLD_GAP 0.1", &
1095 64714 : default_r_val=0.1_dp)
1096 64714 : CALL section_add_keyword(print_key, keyword)
1097 64714 : CALL keyword_release(keyword)
1098 :
1099 : CALL keyword_create(keyword, __LOCATION__, name="STRIDE", &
1100 : description="The stride (X,Y,Z) used to write the cube file "// &
1101 : "(larger values result in smaller cube files). You can provide 3 numbers (for X,Y,Z) or"// &
1102 : " 1 number valid for all components.", &
1103 64714 : usage="STRIDE 2 2 2", n_var=-1, default_i_vals=(/2, 2, 2/), type_of_var=integer_t)
1104 64714 : CALL section_add_keyword(print_key, keyword)
1105 64714 : CALL keyword_release(keyword)
1106 :
1107 64714 : CALL section_add_subsection(section, print_key)
1108 64714 : CALL section_release(print_key)
1109 :
1110 : CALL section_create(gw_dos_section, __LOCATION__, name="GW_DOS", &
1111 : description="Section for printing the spectral function.", &
1112 64714 : n_keywords=6, n_subsections=0, repeats=.FALSE.)
1113 :
1114 : CALL keyword_create(keyword, __LOCATION__, name="LOWER_BOUND", &
1115 : description="Lower bound for GW-DOS in eV.", &
1116 : usage="LOWER_BOUND -20.0", &
1117 : default_r_val=cp_unit_to_cp2k(value=-20.0_dp, unit_str="eV"), &
1118 64714 : unit_str="eV")
1119 64714 : CALL section_add_keyword(gw_dos_section, keyword)
1120 64714 : CALL keyword_release(keyword)
1121 :
1122 : CALL keyword_create(keyword, __LOCATION__, name="UPPER_BOUND", &
1123 : description="Upper bound for GW-DOS in eV.", &
1124 : usage="UPPER_BOUND 5.0", &
1125 : default_r_val=cp_unit_to_cp2k(value=0.0_dp, unit_str="eV"), &
1126 64714 : unit_str="eV")
1127 64714 : CALL section_add_keyword(gw_dos_section, keyword)
1128 64714 : CALL keyword_release(keyword)
1129 :
1130 : CALL keyword_create(keyword, __LOCATION__, name="STEP", &
1131 : description="Difference of two consecutive energy levels for GW-DOS.", &
1132 : usage="STEP 0.1", &
1133 : default_r_val=cp_unit_to_cp2k(value=0.0_dp, unit_str="eV"), &
1134 64714 : unit_str="eV")
1135 64714 : CALL section_add_keyword(gw_dos_section, keyword)
1136 64714 : CALL keyword_release(keyword)
1137 :
1138 : CALL keyword_create(keyword, __LOCATION__, name="MIN_LEVEL_SPECTRAL", &
1139 : description="Lowest energy level to print the self energy to files.", &
1140 : usage="MIN_LEVEL_SPECTRAL 3", &
1141 64714 : default_i_val=1)
1142 64714 : CALL section_add_keyword(gw_dos_section, keyword)
1143 64714 : CALL keyword_release(keyword)
1144 :
1145 : CALL keyword_create(keyword, __LOCATION__, name="MAX_LEVEL_SPECTRAL", &
1146 : description="Highest energy level to print the self energy to files.", &
1147 : usage="MAX_LEVEL_SPECTRAL 6", &
1148 64714 : default_i_val=0)
1149 64714 : CALL section_add_keyword(gw_dos_section, keyword)
1150 64714 : CALL keyword_release(keyword)
1151 :
1152 : CALL keyword_create(keyword, __LOCATION__, name="MIN_LEVEL_SELF_ENERGY", &
1153 : description="Lowest energy level to print the self energy to files.", &
1154 : usage="MIN_LEVEL_SELF_ENERGY 3", &
1155 64714 : default_i_val=1)
1156 64714 : CALL section_add_keyword(gw_dos_section, keyword)
1157 64714 : CALL keyword_release(keyword)
1158 :
1159 : CALL keyword_create(keyword, __LOCATION__, name="MAX_LEVEL_SELF_ENERGY", &
1160 : description="Highest energy level to print the self energy to files.", &
1161 : usage="MAX_LEVEL_SELF_ENERGY 6", &
1162 64714 : default_i_val=0)
1163 64714 : CALL section_add_keyword(gw_dos_section, keyword)
1164 64714 : CALL keyword_release(keyword)
1165 :
1166 : CALL keyword_create(keyword, __LOCATION__, name="BROADENING", &
1167 : description="Broadening parameter for spectral function.", &
1168 : usage="BROADENING 0.001", &
1169 : default_r_val=cp_unit_to_cp2k(value=0.0_dp, unit_str="eV"), &
1170 64714 : unit_str="eV")
1171 64714 : CALL section_add_keyword(gw_dos_section, keyword)
1172 64714 : CALL keyword_release(keyword)
1173 :
1174 64714 : CALL section_add_subsection(section, gw_dos_section)
1175 64714 : CALL section_release(gw_dos_section)
1176 :
1177 64714 : END SUBROUTINE
1178 :
1179 : ! **************************************************************************************************
1180 : !> \brief ...
1181 : !> \param section ...
1182 : ! **************************************************************************************************
1183 64714 : SUBROUTINE create_periodic_gw_correction_section(section)
1184 : TYPE(section_type), POINTER :: section
1185 :
1186 : TYPE(keyword_type), POINTER :: keyword
1187 :
1188 64714 : CPASSERT(.NOT. ASSOCIATED(section))
1189 : CALL section_create(section, __LOCATION__, name="PERIODIC_CORRECTION", &
1190 : description="Parameters influencing correction for periodic GW. Old method, "// &
1191 : "not recommended to use", &
1192 64714 : n_keywords=12, n_subsections=1, repeats=.FALSE.)
1193 :
1194 64714 : NULLIFY (keyword)
1195 :
1196 : CALL keyword_create(keyword, __LOCATION__, name="KPOINTS", &
1197 : description="Specify number of k-points for a single k-point grid. Internally, a "// &
1198 : "Monkhorst-Pack grid is used. Typically, even numbers are chosen such that the Gamma "// &
1199 : "point is excluded from the k-point mesh.", &
1200 : usage="KPOINTS nx ny nz", repeats=.TRUE., &
1201 64714 : n_var=3, type_of_var=integer_t, default_i_vals=(/16, 16, 16/))
1202 64714 : CALL section_add_keyword(section, keyword)
1203 64714 : CALL keyword_release(keyword)
1204 :
1205 : CALL keyword_create(keyword, __LOCATION__, name="NUM_KP_GRIDS", &
1206 : description="Number of k-point grids around the Gamma point with different resolution. "// &
1207 : "E.g. for KPOINTS 4 4 4 and NUM_KP_GRIDS 3, there will be a 3x3x3 Monkhorst-Pack (MP) k-point "// &
1208 : "grid for the whole Brillouin zone (excluding Gamma), another 3x3x3 MP grid with smaller "// &
1209 : "spacing around Gamma (again excluding Gamma) and a very fine 4x4x4 MP grid around Gamma.", &
1210 : usage="NUM_KP_GRIDS 5", &
1211 64714 : default_i_val=1)
1212 64714 : CALL section_add_keyword(section, keyword)
1213 64714 : CALL keyword_release(keyword)
1214 :
1215 : CALL keyword_create(keyword, __LOCATION__, name="EPS_KPOINT", &
1216 : description="If the absolute value of a k-point is below EPS_KPOINT, this kpoint is "// &
1217 : "neglected since the Gamma point is not included in the periodic correction.", &
1218 : usage="EPS_KPOINT 1.0E-4", &
1219 64714 : default_r_val=1.0E-05_dp)
1220 64714 : CALL section_add_keyword(section, keyword)
1221 64714 : CALL keyword_release(keyword)
1222 :
1223 : CALL keyword_create(keyword, __LOCATION__, name="MO_COEFF_GAMMA", &
1224 : description="If true, only the MO coefficients at the Gamma point are used for the periodic "// &
1225 : "correction. Otherwise, the MO coeffs are computed at every k-point which is much more "// &
1226 : "expensive. It should be okay to use the Gamma MO coefficients.", &
1227 : usage="MO_COEFF_GAMMA", &
1228 : default_l_val=.TRUE., &
1229 64714 : lone_keyword_l_val=.TRUE.)
1230 64714 : CALL section_add_keyword(section, keyword)
1231 64714 : CALL keyword_release(keyword)
1232 :
1233 : CALL keyword_create(keyword, __LOCATION__, name="AVERAGE_DEGENERATE_LEVELS", &
1234 : variants=(/"ADL"/), &
1235 : description="If true, the correlation self-energy of degenerate levels is averaged.", &
1236 : usage="AVERAGE_DEGENERATE_LEVELS", &
1237 : default_l_val=.TRUE., &
1238 129428 : lone_keyword_l_val=.TRUE.)
1239 64714 : CALL section_add_keyword(section, keyword)
1240 64714 : CALL keyword_release(keyword)
1241 :
1242 : CALL keyword_create(keyword, __LOCATION__, name="EPS_EIGENVAL", &
1243 : description="Threshold for considering levels as degenerate. Unit: Hartree.", &
1244 : usage="EPS_EIGENVAL 1.0E-5", &
1245 64714 : default_r_val=2.0E-04_dp)
1246 64714 : CALL section_add_keyword(section, keyword)
1247 64714 : CALL keyword_release(keyword)
1248 :
1249 : CALL keyword_create(keyword, __LOCATION__, name="EXTRAPOLATE_KPOINTS", &
1250 : variants=(/"EXTRAPOLATE"/), &
1251 : description="If true, extrapolates the k-point mesh. Only working if k-point mesh numbers are "// &
1252 : "divisible by 4, e.g. 8x8x8 or 12x12x12 is recommended.", &
1253 : usage="EXTRAPOLATE_KPOINTS FALSE", &
1254 : default_l_val=.TRUE., &
1255 129428 : lone_keyword_l_val=.TRUE.)
1256 64714 : CALL section_add_keyword(section, keyword)
1257 64714 : CALL keyword_release(keyword)
1258 :
1259 : CALL keyword_create(keyword, __LOCATION__, name="DO_AUX_BAS_GW", &
1260 : description="If true, use a different basis for the periodic correction. This can be necessary "// &
1261 : "in case a diffused basis is used for GW to converge the HOMO-LUMO gap. In this case, "// &
1262 : "numerical problems may occur due to diffuse functions in the basis. This keyword only works if "// &
1263 : "AUX_GW <basis set> is specified in the kind section for every atom kind.", &
1264 : usage="DO_AUX_BAS_GW TRUE", &
1265 : default_l_val=.FALSE., &
1266 64714 : lone_keyword_l_val=.TRUE.)
1267 64714 : CALL section_add_keyword(section, keyword)
1268 64714 : CALL keyword_release(keyword)
1269 :
1270 : CALL keyword_create(keyword, __LOCATION__, name="FRACTION_AUX_MOS", &
1271 : description="Fraction how many MOs are used in the auxiliary basis.", &
1272 : usage="FRACTION_AUX_MOS 0.6", &
1273 64714 : default_r_val=0.5_dp)
1274 64714 : CALL section_add_keyword(section, keyword)
1275 64714 : CALL keyword_release(keyword)
1276 :
1277 : CALL keyword_create(keyword, __LOCATION__, name="NUM_OMEGA_POINTS", &
1278 : description="Number of Clenshaw-Curtis integration points for the periodic correction in cubic- "// &
1279 : "scaling GW. This variable is a dummy variable for canonical N^4 GW calculations.", &
1280 : usage="NUM_OMEGA_POINTS 200", &
1281 64714 : default_i_val=300)
1282 64714 : CALL section_add_keyword(section, keyword)
1283 64714 : CALL keyword_release(keyword)
1284 :
1285 64714 : END SUBROUTINE
1286 :
1287 : ! **************************************************************************************************
1288 : !> \brief ...
1289 : !> \param section ...
1290 : ! **************************************************************************************************
1291 64714 : SUBROUTINE create_bse_section(section)
1292 : TYPE(section_type), POINTER :: section
1293 :
1294 : TYPE(keyword_type), POINTER :: keyword
1295 : TYPE(section_type), POINTER :: subsection
1296 :
1297 64714 : CPASSERT(.NOT. ASSOCIATED(section))
1298 : CALL section_create(section, __LOCATION__, name="BSE", &
1299 : description="Parameters for a calculation solving the Bethe-Salpeter equation "// &
1300 : "(BSE) for electronic excitations. The full BSE "// &
1301 : "$\left( \begin{array}{cc}A & B\\B & A\end{array} \right)$ "// &
1302 : "$\left( \begin{array}{cc}\mathbf{X}^{(n)}\\\mathbf{Y}^{(n)}\end{array} \right) = "// &
1303 : "\Omega^{(n)}\left(\begin{array}{cc}1&0\\0&-1\end{array}\right)$ "// &
1304 : "$\left(\begin{array}{cc}\mathbf{X}^{(n)}\\\mathbf{Y}^{(n)}\end{array}\right)$ "// &
1305 : "enables, for example, the computation of electronic excitation energies $\Omega^{(n)}$ "// &
1306 : "as well as optical properties. The BSE can be solved by diagonalizing "// &
1307 : "the full ABBA-matrix or by setting B=0, i.e. within the Tamm-Dancoff approximation (TDA). "// &
1308 : "Preliminary reference: Eq. (35) in PRB 92, 045209 (2015); http://dx.doi.org/10.1103/PhysRevB.92.045209", &
1309 64714 : n_keywords=8, n_subsections=3, repeats=.FALSE.)
1310 :
1311 64714 : NULLIFY (keyword)
1312 :
1313 : CALL keyword_create(keyword, __LOCATION__, name="_SECTION_PARAMETERS_", &
1314 : description="Activates BSE calculations.", &
1315 : usage="&BSE .TRUE.", &
1316 64714 : default_l_val=.FALSE., lone_keyword_l_val=.TRUE.)
1317 64714 : CALL section_add_keyword(section, keyword)
1318 64714 : CALL keyword_release(keyword)
1319 :
1320 : CALL keyword_create(keyword, __LOCATION__, name="SPIN_CONFIG", &
1321 : description="Choose between calculation of singlet or triplet excitation (cf. given Reference above).", &
1322 : usage="SPIN_CONFIG TRIPLET", &
1323 : enum_c_vals=s2a("SINGLET", "TRIPLET"), &
1324 : enum_i_vals=(/bse_singlet, bse_triplet/), &
1325 : enum_desc=s2a("Computes singlet excitations.", &
1326 : "Computes triplet excitations."), &
1327 64714 : default_i_val=bse_singlet)
1328 64714 : CALL section_add_keyword(section, keyword)
1329 64714 : CALL keyword_release(keyword)
1330 :
1331 : CALL keyword_create(keyword, __LOCATION__, name="BSE_DIAG_METHOD", &
1332 : description="Method for BSE calculations. "// &
1333 : "Choose between full or iterative diagonalization.", &
1334 : usage="BSE_DIAG_METHOD FULLDIAG", &
1335 : enum_c_vals=s2a("FULLDIAG", "ITERDIAG"), &
1336 : enum_i_vals=(/bse_fulldiag, bse_iterdiag/), &
1337 : enum_desc=s2a("Fully diagonalizes the BSE matrices within the chosen level of approximation.", &
1338 : "Iterative diagonalization has not been implemented yet."), &
1339 64714 : default_i_val=bse_fulldiag)
1340 64714 : CALL section_add_keyword(section, keyword)
1341 64714 : CALL keyword_release(keyword)
1342 :
1343 : CALL keyword_create(keyword, __LOCATION__, name="TDA", &
1344 : description="Level of approximation applied to BSE calculations. "// &
1345 : "Choose between Tamm Dancoff approximation (TDA) and/or diagonalization of the full ABBA-matrix.", &
1346 : usage="TDA ON", &
1347 : enum_c_vals=s2a("ON", "OFF", "TDA+ABBA"), &
1348 : enum_i_vals=(/bse_tda, bse_abba, bse_both/), &
1349 : enum_desc=s2a("The TDA is applied, i.e. B=0.", &
1350 : "The ABBA-matrix is diagonalized, i.e. the TDA is not applied.", &
1351 : "The BSE is solved within the TDA (B=0) as well as for the full ABBA-matrix."), &
1352 64714 : default_i_val=bse_tda)
1353 64714 : CALL section_add_keyword(section, keyword)
1354 64714 : CALL keyword_release(keyword)
1355 :
1356 : CALL keyword_create(keyword, __LOCATION__, name="ENERGY_CUTOFF_OCC", &
1357 : description="Remove all orbitals with indices i,j from A_ia,jb and B_ia,jb with energy difference "// &
1358 : "to HOMO level larger than the given energy cutoff, i.e. "// &
1359 : "$\varepsilon_i\in[\varepsilon_{i=\text{HOMO}}^{GW}-E_\text{cut}^\text{occ},\varepsilon_{i=\text{HOMO}}^{GW}]$. "// &
1360 : "Can be used to accelerate runtime and reduce memory consumption.", &
1361 : usage="ENERGY_CUTOFF_OCC 10.0", unit_str="eV", &
1362 64714 : type_of_var=real_t, default_r_val=-1.0_dp)
1363 64714 : CALL section_add_keyword(section, keyword)
1364 64714 : CALL keyword_release(keyword)
1365 :
1366 : CALL keyword_create(keyword, __LOCATION__, name="ENERGY_CUTOFF_EMPTY", &
1367 : description="Remove all orbitals with indices a,b from A_ia,jb and B_ia,jb with energy difference "// &
1368 : "to LUMO level larger than the given energy cutoff, i.e. "// &
1369 : "$\varepsilon_a\in[\varepsilon_{a=\text{LUMO}}^{GW},\varepsilon_{a=\text{LUMO}}^{GW}+E_\text{cut}^\text{empty}]$. "// &
1370 : "Can be used to accelerate runtime and reduce memory consumption.", &
1371 : usage="ENERGY_CUTOFF_EMPTY 10.0", unit_str="eV", &
1372 64714 : type_of_var=real_t, default_r_val=-1.0_dp)
1373 64714 : CALL section_add_keyword(section, keyword)
1374 64714 : CALL keyword_release(keyword)
1375 :
1376 : CALL keyword_create(keyword, __LOCATION__, name="BSE_DEBUG_PRINT", &
1377 : description="Activates debug print statements in the BSE calculation.", &
1378 : usage="BSE_DEBUG_PRINT .TRUE.", &
1379 64714 : default_l_val=.FALSE., lone_keyword_l_val=.TRUE.)
1380 64714 : CALL section_add_keyword(section, keyword)
1381 64714 : CALL keyword_release(keyword)
1382 :
1383 : CALL keyword_create(keyword, __LOCATION__, name="NUM_PRINT_EXC", &
1384 : description="Number of printed excitation levels with respective "// &
1385 : "energies and oscillator strengths. Does not affect computation time.", &
1386 : usage="NUM_PRINT_EXC 25", &
1387 64714 : default_i_val=25)
1388 64714 : CALL section_add_keyword(section, keyword)
1389 64714 : CALL keyword_release(keyword)
1390 :
1391 : CALL keyword_create(keyword, __LOCATION__, name="NUM_PRINT_EXC_DESCR", &
1392 : description="Number of excitation levels for which the exciton "// &
1393 : "descriptors are computed. Negative or too large "// &
1394 : "NUM_PRINT_EXC_DESCR defaults to NUM_PRINT_EXC.", &
1395 : usage="NUM_PRINT_EXC_DESCR 5", &
1396 64714 : default_i_val=0)
1397 64714 : CALL section_add_keyword(section, keyword)
1398 64714 : CALL keyword_release(keyword)
1399 :
1400 : CALL keyword_create(keyword, __LOCATION__, name="PRINT_DIRECTIONAL_EXC_DESCR", &
1401 : description="Activates printing of exciton descriptors per direction.", &
1402 : usage="PRINT_DIRECTIONAL_EXC_DESCR .TRUE.", &
1403 64714 : default_l_val=.FALSE., lone_keyword_l_val=.TRUE.)
1404 64714 : CALL section_add_keyword(section, keyword)
1405 64714 : CALL keyword_release(keyword)
1406 :
1407 : CALL keyword_create(keyword, __LOCATION__, name="EPS_X", &
1408 : description="Threshold for printing contributions of singleparticle "// &
1409 : "transitions, i.e. elements of the eigenvectors $X_{ia}^{(n)}$ and $Y_{ia}^{(n)}$.", &
1410 : usage="EPS_X 0.1", &
1411 64714 : type_of_var=real_t, default_r_val=0.1_dp)
1412 64714 : CALL section_add_keyword(section, keyword)
1413 64714 : CALL keyword_release(keyword)
1414 :
1415 : CALL keyword_create(keyword, __LOCATION__, name="USE_KS_ENERGIES", &
1416 : description="Uses KS energies instead of GW quasiparticle energies.", &
1417 : usage="USE_KS_ENERGIES .TRUE.", &
1418 64714 : default_l_val=.FALSE., lone_keyword_l_val=.TRUE.)
1419 64714 : CALL section_add_keyword(section, keyword)
1420 64714 : CALL keyword_release(keyword)
1421 :
1422 64714 : NULLIFY (subsection)
1423 64714 : CALL create_bse_screening_section(subsection)
1424 64714 : CALL section_add_subsection(section, subsection)
1425 64714 : CALL section_release(subsection)
1426 :
1427 64714 : NULLIFY (subsection)
1428 64714 : CALL create_bse_iterat_section(subsection)
1429 64714 : CALL section_add_subsection(section, subsection)
1430 64714 : CALL section_release(subsection)
1431 :
1432 64714 : NULLIFY (subsection)
1433 64714 : CALL create_bse_spectrum_section(subsection)
1434 64714 : CALL section_add_subsection(section, subsection)
1435 64714 : CALL section_release(subsection)
1436 :
1437 64714 : NULLIFY (subsection)
1438 64714 : CALL create_bse_nto_section(subsection)
1439 64714 : CALL section_add_subsection(section, subsection)
1440 64714 : CALL section_release(subsection)
1441 :
1442 64714 : END SUBROUTINE
1443 :
1444 64714 : SUBROUTINE create_bse_screening_section(section)
1445 : TYPE(section_type), POINTER :: section
1446 :
1447 : TYPE(keyword_type), POINTER :: keyword
1448 :
1449 64714 : CPASSERT(.NOT. ASSOCIATED(section))
1450 :
1451 : CALL section_create(section, __LOCATION__, name="SCREENING_IN_W", &
1452 : description="Screening $\epsilon$ applied to $W(\omega=0)=\epsilon^{-1}(\omega=0) v $ "// &
1453 : "in the BSE calculation. Besides default BSE, i.e. $W_0$ (screening with DFT energies), "// &
1454 : "a fixed $\alpha = \epsilon^{-1}(\omega)$ can be applied, which is similar to the mixing "// &
1455 : "parameter for hybrid functionals in LR-TDDFT. In addition, the keywords TDHF "// &
1456 : "(no screening - $\alpha = 1$) and RPA (infinite screening - $\alpha = 0$) can be applied.", &
1457 64714 : n_keywords=2, n_subsections=0, repeats=.FALSE.)
1458 :
1459 64714 : NULLIFY (keyword)
1460 :
1461 : CALL keyword_create( &
1462 : keyword, __LOCATION__, name="_SECTION_PARAMETERS_", &
1463 : description="Shortcut for the most common functional combinations.", &
1464 : usage="&xc_functional BLYP", &
1465 : enum_c_vals=s2a("W_0", "TDHF", "RPA", "ALPHA"), &
1466 : enum_i_vals=(/bse_screening_w0, bse_screening_tdhf, bse_screening_rpa, bse_screening_alpha/), &
1467 : enum_desc=s2a("The Coulomb interaction is screened by applying DFT energies "// &
1468 : "$\varepsilon_p^{DFT}$, which is typically used for GW-BSE and "// &
1469 : "often labeled as $W_0$.", &
1470 : "The Coulomb interaction is not screened, i.e. $W_{pq,rs}(\omega=0) "// &
1471 : "\rightarrow v_{pq,rs}$ enters.", &
1472 : "Infinite screening is applied, i.e. $W_{pq,rs}(\omega=0) \rightarrow 0$.", &
1473 : "Arbitrary screening parameter. Specify within section."), &
1474 : default_i_val=bse_screening_w0, &
1475 64714 : lone_keyword_i_val=bse_screening_w0)
1476 64714 : CALL section_add_keyword(section, keyword)
1477 64714 : CALL keyword_release(keyword)
1478 :
1479 : CALL keyword_create(keyword, __LOCATION__, name="ALPHA", &
1480 : description="Screening parameter similar to the mixing in hybrid functionals used in TDDFT. "// &
1481 : "$\alpha$ mimicks the screening $\epsilon^{-1}(\omega)$ and enforces $W = \alpha v$ "// &
1482 : "in the BSE calculation.", &
1483 : usage="ALPHA 0.25", &
1484 64714 : type_of_var=real_t, default_r_val=-1.00_dp)
1485 64714 : CALL section_add_keyword(section, keyword)
1486 64714 : CALL keyword_release(keyword)
1487 :
1488 64714 : END SUBROUTINE create_bse_screening_section
1489 :
1490 64714 : SUBROUTINE create_bse_nto_section(print_key)
1491 : TYPE(section_type), POINTER :: print_key
1492 :
1493 : TYPE(keyword_type), POINTER :: keyword
1494 :
1495 64714 : CPASSERT(.NOT. ASSOCIATED(print_key))
1496 :
1497 : CALL cp_print_key_section_create(print_key, __LOCATION__, name="NTO_ANALYSIS", &
1498 : description="Perform a natural transition orbital analysis, i.e. the transition density matrix "// &
1499 : "$T^{(n)}=\left( \begin{array}{cc}0& {X}^{(n)}\\ \left({Y}^{(n)} \right)^T & 0\end{array} \right)$ "// &
1500 : "is decomposed into its singular values "// &
1501 : "$T^{(n)} = {U}^{(n)} {\Lambda^{(n)}} \left({V}^{(n)}\right)^T$ "// &
1502 : "in order to compute the NTO pairs "// &
1503 : "$\phi_I^{(n)}(\mathbf{r}_e) = \sum_{p=1}^{N_b} \varphi_p(\mathbf{r}_e) V_{p,I}^{(n)}$ for the electron and "// &
1504 : "$\chi_I^{(n)}(\mathbf{r}_h) = \sum_{q=1}^{N_b} \varphi_q(\mathbf{r}_h) U_{q,I}^{(n)}$ for the hole.", &
1505 : print_level=debug_print_level + 1, & ! Corresponds to "off" as default behavior
1506 64714 : filename="BSE-NTO_ANALYSIS") ! All other print levels will trigger the analysis
1507 : ! cf. input/cp_output_handling.F:cp_print_key_section_create
1508 :
1509 64714 : NULLIFY (keyword)
1510 : CALL keyword_create(keyword, __LOCATION__, name="EPS_NTO_EIGVAL", &
1511 : description="Threshold for NTO eigenvalues, i.e. only "// &
1512 : "${\left(\lambda_I^{(n)}\right)}^2$ > EPS_NTO_EIGVAL are considered.", &
1513 : usage="EPS_NTO_EIGVAL 0.01", &
1514 : n_var=1, &
1515 : type_of_var=real_t, &
1516 64714 : default_r_val=0.01_dp)
1517 64714 : CALL section_add_keyword(print_key, keyword)
1518 64714 : CALL keyword_release(keyword)
1519 :
1520 : CALL keyword_create(keyword, __LOCATION__, name="EPS_OSC_STR", &
1521 : description="Threshold of oscillator strengths $f^{(n)}$ for an excitation level. "// &
1522 : "In case, the excitation level n has a smaller oscillator strength, the "// &
1523 : "NTOs for this excitation level are not printed.", &
1524 : usage="EPS_OSC_STR 0.01", &
1525 : n_var=1, &
1526 : type_of_var=real_t, &
1527 64714 : default_r_val=-1.0_dp)
1528 64714 : CALL section_add_keyword(print_key, keyword)
1529 64714 : CALL keyword_release(keyword)
1530 :
1531 : CALL keyword_create(keyword, __LOCATION__, name="NUM_PRINT_EXC_NTOS", &
1532 : description="Number of excitation level $n$ up to which NTOs are printed. "// &
1533 : "By default, this is set to NUM_PRINT_EXC. Negative or too large "// &
1534 : "NUM_PRINT_EXC_NTOS defaults to NUM_PRINT_EXC.", &
1535 : usage="NUM_PRINT_EXC_NTOS 5", &
1536 : n_var=1, &
1537 : type_of_var=integer_t, &
1538 64714 : default_i_val=-1)
1539 64714 : CALL section_add_keyword(print_key, keyword)
1540 64714 : CALL keyword_release(keyword)
1541 :
1542 : CALL keyword_create(keyword, __LOCATION__, name="STATE_LIST", &
1543 : description="Specifies a list of excitation levels $n$ for which NTOs are printed. "// &
1544 : "Overrides NUM_PRINT_EXC_NTOS.", &
1545 : usage="STATE_LIST {integer} {integer} .. {integer}", &
1546 64714 : n_var=-1, type_of_var=integer_t)
1547 64714 : CALL section_add_keyword(print_key, keyword)
1548 64714 : CALL keyword_release(keyword)
1549 :
1550 : CALL keyword_create(keyword, __LOCATION__, name="CUBE_FILES", &
1551 : description="Print NTOs on Cube Files", &
1552 : usage="CUBE_FILES {logical}", repeats=.FALSE., n_var=1, &
1553 64714 : default_l_val=.TRUE., lone_keyword_l_val=.TRUE., type_of_var=logical_t)
1554 64714 : CALL section_add_keyword(print_key, keyword)
1555 64714 : CALL keyword_release(keyword)
1556 :
1557 : CALL keyword_create(keyword, __LOCATION__, name="STRIDE", &
1558 : description="The stride (X,Y,Z) used to write the cube file "// &
1559 : "(larger values result in smaller cube files). Provide 3 numbers (for X,Y,Z) or"// &
1560 : " 1 number valid for all components.", &
1561 64714 : usage="STRIDE 2 2 2", n_var=-1, default_i_vals=(/2, 2, 2/), type_of_var=integer_t)
1562 64714 : CALL section_add_keyword(print_key, keyword)
1563 64714 : CALL keyword_release(keyword)
1564 :
1565 : CALL keyword_create(keyword, __LOCATION__, name="APPEND", &
1566 : description="append the cube files when they already exist", &
1567 64714 : default_l_val=.FALSE., lone_keyword_l_val=.TRUE.)
1568 64714 : CALL section_add_keyword(print_key, keyword)
1569 64714 : CALL keyword_release(keyword)
1570 :
1571 64714 : END SUBROUTINE create_bse_nto_section
1572 :
1573 64714 : SUBROUTINE create_bse_spectrum_section(section)
1574 : TYPE(section_type), POINTER :: section
1575 :
1576 : TYPE(keyword_type), POINTER :: keyword
1577 :
1578 64714 : CPASSERT(.NOT. ASSOCIATED(section))
1579 : CALL section_create(section, __LOCATION__, name="BSE_SPECTRUM", &
1580 : description="Parameters influencing the output of the optical absorption spectrum, i.e. "// &
1581 : "the dynamical dipole polarizability tensor $\alpha_{\mu,\mu'}(\omega)$ "// &
1582 : "($\mu,\mu'\in\{x,y,z\}$), obtained from a BSE calculation, which is defined as "// &
1583 : "$ \alpha_{\mu,\mu'}(\omega) = \sum_n \frac{2 E^{(n)} d^{(n)}_{\mu} d^{(n)}_{\mu'}} "// &
1584 : "{(\omega+i\eta)^2-\left(\Omega^{(n)}\right)^2} $. "// &
1585 : "The printed file will contain the specified frequency range $\omega$ and the "// &
1586 : "corresponding imaginary part of the average $\bar{\alpha}(\omega)=\frac{1}{3}\mathrm{Tr} "// &
1587 : "\left[ \alpha_{\mu,\mu'}(\omega)\right]$ as well as of the elements of $\alpha_{\mu,\mu'}(\omega)$.", &
1588 64714 : n_keywords=9, n_subsections=0, repeats=.FALSE.)
1589 :
1590 64714 : NULLIFY (keyword)
1591 :
1592 : CALL keyword_create(keyword, __LOCATION__, name="_SECTION_PARAMETERS_", &
1593 : description="Activates printing of optical absorption spectrum from the BSE calculation.", &
1594 : usage="&BSE .TRUE.", &
1595 64714 : default_l_val=.FALSE., lone_keyword_l_val=.TRUE.)
1596 64714 : CALL section_add_keyword(section, keyword)
1597 64714 : CALL keyword_release(keyword)
1598 :
1599 : CALL keyword_create(keyword, __LOCATION__, name="FREQUENCY_STEP_SIZE", &
1600 : description="Step size of frequencies for the optical absorption spectrum.", &
1601 : usage="FREQUENCY_STEP_SIZE 0.1", unit_str="eV", &
1602 : type_of_var=real_t, &
1603 64714 : default_r_val=cp_unit_to_cp2k(value=0.1_dp, unit_str="eV"))
1604 64714 : CALL section_add_keyword(section, keyword)
1605 64714 : CALL keyword_release(keyword)
1606 :
1607 : CALL keyword_create(keyword, __LOCATION__, name="FREQUENCY_STARTING_POINT", &
1608 : description="First frequency to print in the optical absorption spectrum.", &
1609 : usage="FREQUENCY_STARTING_POINT 0", unit_str="eV", &
1610 64714 : type_of_var=real_t, default_r_val=0.0_dp)
1611 64714 : CALL section_add_keyword(section, keyword)
1612 64714 : CALL keyword_release(keyword)
1613 :
1614 : CALL keyword_create(keyword, __LOCATION__, name="FREQUENCY_END_POINT", &
1615 : description="Last frequency to print in the optical absorption spectrum.", &
1616 : usage="FREQUENCY_END_POINT 0", unit_str="eV", &
1617 : type_of_var=real_t, &
1618 64714 : default_r_val=cp_unit_to_cp2k(value=100.0_dp, unit_str="eV"))
1619 64714 : CALL section_add_keyword(section, keyword)
1620 64714 : CALL keyword_release(keyword)
1621 :
1622 : CALL keyword_create(keyword, __LOCATION__, name="ETA_LIST", &
1623 : description="List of broadening of the peaks in the optical absorption spectrum.", &
1624 : usage="ETA_LIST 0.01 ...", unit_str="eV", &
1625 : default_r_vals=(/cp_unit_to_cp2k(value=0.01_dp, unit_str="eV")/), &
1626 129428 : type_of_var=real_t, n_var=-1)
1627 64714 : CALL section_add_keyword(section, keyword)
1628 64714 : CALL keyword_release(keyword)
1629 :
1630 64714 : END SUBROUTINE
1631 :
1632 : ! **************************************************************************************************
1633 : !> \brief ...
1634 : !> \param section ...
1635 : ! **************************************************************************************************
1636 64714 : SUBROUTINE create_bse_iterat_section(section)
1637 : TYPE(section_type), POINTER :: section
1638 :
1639 : TYPE(keyword_type), POINTER :: keyword
1640 :
1641 64714 : CPASSERT(.NOT. ASSOCIATED(section))
1642 : CALL section_create(section, __LOCATION__, name="BSE_ITERAT", &
1643 : description="Parameters influencing the iterative Bethe-Salpeter calculation. "// &
1644 : "The iterative solver has not been fully implemented yet.", &
1645 64714 : n_keywords=9, n_subsections=0, repeats=.FALSE.)
1646 :
1647 64714 : NULLIFY (keyword)
1648 :
1649 : CALL keyword_create(keyword, __LOCATION__, name="DAVIDSON_ABORT_COND", &
1650 : description="Desired abortion condition for Davidson solver", &
1651 : usage="DAVIDSON_ABORT_COND OR", &
1652 : enum_c_vals=s2a("EN", "RES", "OR"), &
1653 : enum_i_vals=(/bse_iter_en_cond, bse_iter_res_cond, bse_iter_both_cond/), &
1654 : enum_desc=s2a("Uses energy threshold for successfully exiting solver.", &
1655 : "Uses residual threshold for successfully exiting solver.", &
1656 : "Uses either energy or residual threshold for successfully exiting solver."), &
1657 64714 : default_i_val=bse_iter_en_cond)
1658 64714 : CALL section_add_keyword(section, keyword)
1659 64714 : CALL keyword_release(keyword)
1660 :
1661 : CALL keyword_create(keyword, __LOCATION__, name="NUM_EXC_EN", &
1662 : description="Number of lowest excitation energies to be computed.", &
1663 : usage="NUM_EXC_EN 3", &
1664 64714 : default_i_val=3)
1665 64714 : CALL section_add_keyword(section, keyword)
1666 64714 : CALL keyword_release(keyword)
1667 :
1668 : CALL keyword_create(keyword, __LOCATION__, name="NUM_ADD_START_Z_SPACE", &
1669 : description="Determines the initial dimension of the subspace as "// &
1670 : "dim = (NUM_EXC_EN+NUM_ADD_START_Z_SPACE)", &
1671 : usage="NUM_ADD_START_Z_SPACE 1", &
1672 64714 : default_i_val=0)
1673 64714 : CALL section_add_keyword(section, keyword)
1674 64714 : CALL keyword_release(keyword)
1675 :
1676 : CALL keyword_create(keyword, __LOCATION__, name="FAC_MAX_Z_SPACE", &
1677 : description="Factor to determine maximum dimension of the Davidson subspace. "// &
1678 : "dimension = (NUM_EXC_EN+NUM_ADD_START_Z_SPACE)*FAC_MAX_Z_SPACE", &
1679 : usage="FAC_MAX_Z_SPACE 5", &
1680 64714 : default_i_val=5)
1681 64714 : CALL section_add_keyword(section, keyword)
1682 64714 : CALL keyword_release(keyword)
1683 :
1684 : CALL keyword_create(keyword, __LOCATION__, name="NUM_NEW_T", &
1685 : description="Number of new t vectors added. "// &
1686 : "Must be smaller/equals (NUM_EXC_EN+NUM_ADD_START_Z_SPACE)", &
1687 : usage="NUM_NEW_T 4", &
1688 64714 : default_i_val=1)
1689 64714 : CALL section_add_keyword(section, keyword)
1690 64714 : CALL keyword_release(keyword)
1691 :
1692 : CALL keyword_create(keyword, __LOCATION__, name="EPS_RES", &
1693 : description="Threshold for stopping the iteration for computing the transition energies. "// &
1694 : "If the residuals inside the Davidson space change by less than EPS_RES (in eV), the iteration "// &
1695 : "stops.", &
1696 : usage="EPS_RES 0.001", unit_str="eV", &
1697 64714 : type_of_var=real_t, default_r_val=0.001_dp)
1698 64714 : CALL section_add_keyword(section, keyword)
1699 64714 : CALL keyword_release(keyword)
1700 :
1701 : CALL keyword_create(keyword, __LOCATION__, name="EPS_EXC_EN", &
1702 : description="Threshold for stopping the iteration for computing the transition energies. "// &
1703 : "If the desired excitation energies change by less than EPS_EXC_EN (in eV), the iteration "// &
1704 : "stops.", &
1705 : usage="EPS_EXC_EN 0.001", unit_str="eV", &
1706 64714 : type_of_var=real_t, default_r_val=0.001_dp)
1707 64714 : CALL section_add_keyword(section, keyword)
1708 64714 : CALL keyword_release(keyword)
1709 :
1710 : CALL keyword_create(keyword, __LOCATION__, name="NUM_DAVIDSON_ITER", &
1711 : description="Maximum number of iterations for determining the transition energies.", &
1712 : usage="NUM_DAVIDSON_ITER 100", &
1713 64714 : default_i_val=100)
1714 64714 : CALL section_add_keyword(section, keyword)
1715 64714 : CALL keyword_release(keyword)
1716 :
1717 : CALL keyword_create(keyword, __LOCATION__, name="Z_SPACE_ENERGY_CUTOFF", &
1718 : description="Cutoff (in eV) for maximal energy difference entering the A matrix. "// &
1719 : "Per default and for negative values, there is no cutoff applied.", &
1720 : usage="Z_SPACE_ENERGY_CUTOFF 60", unit_str="eV", &
1721 64714 : type_of_var=real_t, default_r_val=-1.0_dp)
1722 64714 : CALL section_add_keyword(section, keyword)
1723 64714 : CALL keyword_release(keyword)
1724 64714 : END SUBROUTINE
1725 :
1726 : ! **************************************************************************************************
1727 : !> \brief ...
1728 : !> \param section ...
1729 : ! **************************************************************************************************
1730 64714 : SUBROUTINE create_ic_section(section)
1731 : TYPE(section_type), POINTER :: section
1732 :
1733 : TYPE(keyword_type), POINTER :: keyword
1734 :
1735 64714 : CPASSERT(.NOT. ASSOCIATED(section))
1736 : CALL section_create(section, __LOCATION__, name="IC", &
1737 : description="Parameters influencing the image charge correction. "// &
1738 : "The image plane is always an xy plane, so adjust the molecule according "// &
1739 : "to that. ", &
1740 64714 : n_keywords=3, n_subsections=1, repeats=.FALSE.)
1741 :
1742 64714 : NULLIFY (keyword)
1743 :
1744 : CALL keyword_create(keyword, __LOCATION__, name="PRINT_IC_LIST", &
1745 : description="If true, the image charge correction values are printed in a list, "// &
1746 : "such that it can be used as input for a subsequent evGW calculation.", &
1747 : usage="PRINT_IC_LIST .TRUE.", &
1748 : default_l_val=.FALSE., &
1749 64714 : lone_keyword_l_val=.TRUE.)
1750 64714 : CALL section_add_keyword(section, keyword)
1751 64714 : CALL keyword_release(keyword)
1752 :
1753 : CALL keyword_create(keyword, __LOCATION__, name="EPS_DIST", &
1754 : description="Threshold where molecule and image molecule have to coincide. ", &
1755 : usage="EPS_DIST 0.1", unit_str="angstrom", &
1756 64714 : type_of_var=real_t, default_r_val=3.0E-02_dp, repeats=.FALSE.)
1757 64714 : CALL section_add_keyword(section, keyword)
1758 64714 : CALL keyword_release(keyword)
1759 :
1760 64714 : END SUBROUTINE
1761 :
1762 : ! **************************************************************************************************
1763 : !> \brief ...
1764 : !> \param section ...
1765 : ! **************************************************************************************************
1766 64714 : SUBROUTINE create_low_scaling(section)
1767 : TYPE(section_type), POINTER :: section
1768 :
1769 : TYPE(keyword_type), POINTER :: keyword
1770 : TYPE(section_type), POINTER :: subsection
1771 :
1772 64714 : CPASSERT(.NOT. ASSOCIATED(section))
1773 : CALL section_create( &
1774 : section, __LOCATION__, name="LOW_SCALING", &
1775 : description="Cubic scaling RI-RPA, GW and Laplace-SOS-MP2 method using the imaginary time formalism. "// &
1776 : "EPS_GRID in WFC_GPW section controls accuracy / req. memory for 3-center integrals. "// &
1777 : "SORT_BASIS EXP should be specified in DFT section.", &
1778 : n_keywords=12, n_subsections=2, repeats=.FALSE., &
1779 258856 : citations=(/Wilhelm2016b, Wilhelm2018, Bussy2023/))
1780 :
1781 64714 : NULLIFY (keyword)
1782 : CALL keyword_create(keyword, __LOCATION__, name="_SECTION_PARAMETERS_", &
1783 : description="Activates cubic-scaling RPA, GW and Laplace-SOS-MP2 calculations.", &
1784 : usage="&LOW_SCALING .TRUE.", &
1785 64714 : default_l_val=.FALSE., lone_keyword_l_val=.TRUE.)
1786 64714 : CALL section_add_keyword(section, keyword)
1787 64714 : CALL keyword_release(keyword)
1788 :
1789 : CALL keyword_create(keyword, __LOCATION__, name="MEMORY_CUT", &
1790 : description="Reduces memory for sparse tensor contractions by this factor. "// &
1791 : "A high value leads to some loss of performance. "// &
1792 : "This memory reduction factor applies to storage of the tensors 'M occ' / 'M virt' "// &
1793 : "but does not reduce storage of '3c ints'.", &
1794 : usage="MEMORY_CUT 16", &
1795 64714 : default_i_val=5)
1796 64714 : CALL section_add_keyword(section, keyword)
1797 64714 : CALL keyword_release(keyword)
1798 :
1799 : CALL keyword_create(keyword, __LOCATION__, name="MEMORY_INFO", &
1800 : description="Decide whether to print memory info on the sparse matrices.", &
1801 : usage="MEMORY_INFO", &
1802 : default_l_val=.FALSE., &
1803 64714 : lone_keyword_l_val=.TRUE.)
1804 64714 : CALL section_add_keyword(section, keyword)
1805 64714 : CALL keyword_release(keyword)
1806 :
1807 : CALL keyword_create( &
1808 : keyword, __LOCATION__, name="EPS_FILTER", &
1809 : description="Determines a threshold for the DBCSR based multiply. "// &
1810 : "Normally, this EPS_FILTER determines accuracy and timing of low-scaling RPA and GW calculations.", &
1811 : usage="EPS_FILTER 1.0E-10 ", type_of_var=real_t, &
1812 64714 : default_r_val=1.0E-9_dp)
1813 64714 : CALL section_add_keyword(section, keyword)
1814 64714 : CALL keyword_release(keyword)
1815 :
1816 : CALL keyword_create( &
1817 : keyword, __LOCATION__, name="EPS_FILTER_FACTOR", &
1818 : description="Multiply EPS_FILTER with this factor to determine filter epsilon "// &
1819 : "for DBCSR based multiply P(it)=(Mocc(it))^T*Mvirt(it) "// &
1820 : "Default should be kept.", &
1821 : type_of_var=real_t, &
1822 64714 : default_r_val=10.0_dp)
1823 64714 : CALL section_add_keyword(section, keyword)
1824 64714 : CALL keyword_release(keyword)
1825 :
1826 : CALL keyword_create( &
1827 : keyword, __LOCATION__, &
1828 : name="EPS_STORAGE_SCALING", &
1829 : variants=(/"EPS_STORAGE"/), &
1830 : description="Scaling factor to scale EPS_FILTER. Storage threshold for compression "// &
1831 : "will be EPS_FILTER*EPS_STORAGE_SCALING.", &
1832 129428 : default_r_val=1.0E-3_dp)
1833 64714 : CALL section_add_keyword(section, keyword)
1834 64714 : CALL keyword_release(keyword)
1835 :
1836 : CALL keyword_create( &
1837 : keyword, __LOCATION__, &
1838 : name="DO_KPOINTS", &
1839 : description="Besides in DFT, this keyword has to be switched on if one wants to do kpoints in. "// &
1840 : "cubic RPA.", &
1841 : usage="DO_KPOINTS", &
1842 : default_l_val=.FALSE., &
1843 64714 : lone_keyword_l_val=.TRUE.)
1844 64714 : CALL section_add_keyword(section, keyword)
1845 64714 : CALL keyword_release(keyword)
1846 :
1847 : CALL keyword_create( &
1848 : keyword, __LOCATION__, name="KPOINTS", &
1849 : description="Keyword activates periodic, low-scaling GW calculations (&LOW_SCALING section also needed). "// &
1850 : "For periodic calculations, kpoints are used for the density response, the "// &
1851 : "Coulomb interaction and the screened Coulomb interaction. For 2d periodic systems, e.g. xz "// &
1852 : "periodicity, please also specify KPOINTS, e.g. N_x 1 N_z.", &
1853 : usage="KPOINTS N_x N_y N_z", &
1854 64714 : n_var=3, type_of_var=integer_t, default_i_vals=(/0, 0, 0/))
1855 64714 : CALL section_add_keyword(section, keyword)
1856 64714 : CALL keyword_release(keyword)
1857 :
1858 : CALL keyword_create( &
1859 : keyword, __LOCATION__, &
1860 : name="KPOINT_WEIGHTS_W", &
1861 : description="For kpoints in low-scaling GW, a Monkhorst-Pack mesh is used. The screened Coulomb "// &
1862 : "interaction W(k) needs special care near the Gamma point (e.g. in 3d, W(k) diverges at the "// &
1863 : "Gamma point with W(k) ~ k^alpha). KPOINT_WEIGHTS_W decides how the weights of the "// &
1864 : "Monkhorst-Pack mesh are chosen to compute W(R) = int_BZ W(k) exp(ikR) dk (BZ=Brllouin zone). ", &
1865 : usage="KPOINT_WEIGHTS_W AUTO", &
1866 : enum_c_vals=s2a("TAILORED", "AUTO", "UNIFORM"), &
1867 : enum_i_vals=(/kp_weights_W_tailored, kp_weights_W_auto, kp_weights_W_uniform/), &
1868 : enum_desc=s2a("Choose k-point integration weights such that the function f(k)=k^alpha is "// &
1869 : "exactly integrated. alpha is specified using EXPONENT_TAILORED_WEIGHTS.", &
1870 : "As 'TAILORED', but alpha is chosen automatically according to dimensionality "// &
1871 : "(3D: alpha = -2 for 3D, 2D: alpha = -1 for exchange self-energy, uniform "// &
1872 : "weights for correlation self-energy).", &
1873 : "Choose the same weight for every k-point (original Monkhorst-Pack method)."), &
1874 64714 : default_i_val=kp_weights_W_uniform)
1875 64714 : CALL section_add_keyword(section, keyword)
1876 64714 : CALL keyword_release(keyword)
1877 :
1878 : CALL keyword_create( &
1879 : keyword, __LOCATION__, &
1880 : name="EXPONENT_TAILORED_WEIGHTS", &
1881 : description="Gives the exponent of exactly integrated function in case 'KPOINT_WEIGHTS_W "// &
1882 : "TAILORED' is chosen.", &
1883 : usage="EXPONENT_TAILORED_WEIGHTS -2", &
1884 64714 : default_r_val=-2.0_dp)
1885 64714 : CALL section_add_keyword(section, keyword)
1886 64714 : CALL keyword_release(keyword)
1887 :
1888 : CALL keyword_create( &
1889 : keyword, __LOCATION__, &
1890 : name="REGULARIZATION_RI", &
1891 : description="Parameter to reduce the expansion coefficients in RI for periodic GW. Larger parameter "// &
1892 : "means smaller expansion coefficients that leads to a more stable calculation at the price "// &
1893 : "of a slightly worse RI approximation. In case the parameter 0.0 is chosen, ordinary RI is used.", &
1894 : usage="REGULARIZATION_RI 1.0E-4", &
1895 64714 : default_r_val=0.0_dp)
1896 64714 : CALL section_add_keyword(section, keyword)
1897 64714 : CALL keyword_release(keyword)
1898 :
1899 : CALL keyword_create( &
1900 : keyword, __LOCATION__, &
1901 : name="EPS_EIGVAL_S", &
1902 : description="Parameter to reduce the expansion coefficients in RI for periodic GW. Removes all "// &
1903 : "eigenvectors and eigenvalues of S_PQ(k) that are smaller than EPS_EIGVAL_S. ", &
1904 : usage="EPS_EIGVAL_S 1.0E-3", &
1905 64714 : default_r_val=0.0_dp)
1906 64714 : CALL section_add_keyword(section, keyword)
1907 64714 : CALL keyword_release(keyword)
1908 :
1909 : CALL keyword_create( &
1910 : keyword, __LOCATION__, &
1911 : name="EPS_EIGVAL_S_GAMMA", &
1912 : description="Parameter to reduce the expansion coefficients in RI for periodic GW. Removes all "// &
1913 : "eigenvectors and eigenvalues of M_PQ(k=0) that are smaller than EPS_EIGVAL_S. ", &
1914 : usage="EPS_EIGVAL_S_GAMMA 1.0E-3", &
1915 64714 : default_r_val=0.0_dp)
1916 64714 : CALL section_add_keyword(section, keyword)
1917 64714 : CALL keyword_release(keyword)
1918 :
1919 : CALL keyword_create( &
1920 : keyword, __LOCATION__, &
1921 : name="MAKE_CHI_POS_DEFINITE", &
1922 : description="If true, makes eigenvalue decomposition of chi(iw,k) and removes negative "// &
1923 : "eigenvalues. May increase computational cost significantly. Only recommended to try in case "// &
1924 : "Cholesky decomposition of epsilon(iw,k) fails.", &
1925 : usage="MAKE_CHI_POS_DEFINITE", &
1926 : default_l_val=.TRUE., &
1927 64714 : lone_keyword_l_val=.TRUE.)
1928 64714 : CALL section_add_keyword(section, keyword)
1929 64714 : CALL keyword_release(keyword)
1930 :
1931 : CALL keyword_create( &
1932 : keyword, __LOCATION__, &
1933 : name="MAKE_OVERLAP_MAT_AO_POS_DEFINITE", &
1934 : description="If true, makes eigenvalue decomposition of S_mu,nu(k) and removes negative "// &
1935 : "eigenvalues. Slightly increases computational cost. Only recommended to try in case "// &
1936 : "Cholesky decomposition of S_mu,nu(k) fails (error message: Cholesky decompose failed: "// &
1937 : "matrix is not positive definite or ill-conditioned; when calling create_kp_and_calc_kp_orbitals).", &
1938 : usage="MAKE_OVERLAP_MAT_AO_POS_DEFINITE", &
1939 : default_l_val=.FALSE., &
1940 64714 : lone_keyword_l_val=.TRUE.)
1941 64714 : CALL section_add_keyword(section, keyword)
1942 64714 : CALL keyword_release(keyword)
1943 :
1944 : CALL keyword_create( &
1945 : keyword, __LOCATION__, &
1946 : name="DO_EXTRAPOLATE_KPOINTS", &
1947 : description="If true, use a larger k-mesh to extrapolate the k-point integration of W. "// &
1948 : "For example, in 2D, when using KPOINTS 4 4 1, an additional 6x6x1 mesh will be used to "// &
1949 : "extrapolate the k-point integration of W with N_k^-0.5, where Nk is the number of k-points.", &
1950 : usage="DO_EXTRAPOLATE_KPOINTS FALSE", &
1951 : default_l_val=.TRUE., &
1952 64714 : lone_keyword_l_val=.TRUE.)
1953 64714 : CALL section_add_keyword(section, keyword)
1954 64714 : CALL keyword_release(keyword)
1955 :
1956 : CALL keyword_create( &
1957 : keyword, __LOCATION__, &
1958 : name="TRUNC_COULOMB_RI_X", &
1959 : description="If true, use the truncated Coulomb operator for the exchange-self-energy in "// &
1960 : "periodic GW.", &
1961 : usage="TRUNC_COULOMB_RI_X", &
1962 : default_l_val=.TRUE., &
1963 64714 : lone_keyword_l_val=.TRUE.)
1964 64714 : CALL section_add_keyword(section, keyword)
1965 64714 : CALL keyword_release(keyword)
1966 :
1967 : CALL keyword_create( &
1968 : keyword, __LOCATION__, &
1969 : name="REL_CUTOFF_TRUNC_COULOMB_RI_X", &
1970 : description="Only active in case TRUNC_COULOMB_RI_X = True. Normally, relative cutoff = 0.5 is "// &
1971 : "good choice; still needs to be evaluated for RI schemes. ", &
1972 : usage="REL_CUTOFF_TRUNC_COULOMB_RI_X 0.3", &
1973 64714 : default_r_val=0.5_dp)
1974 64714 : CALL section_add_keyword(section, keyword)
1975 64714 : CALL keyword_release(keyword)
1976 :
1977 : CALL keyword_create( &
1978 : keyword, __LOCATION__, &
1979 : name="KEEP_QUADRATURE", &
1980 : variants=s2a("KEEP_WEIGHTS", "KEEP_QUAD", "KEEP_WEIGHT"), &
1981 : description="Keep the Laplace quadrature defined at the first energy evaluations throughout "// &
1982 : "the run. Allows to have consistent force evaluations.", &
1983 : usage="KEEP_QUADRATURE", &
1984 : default_l_val=.TRUE., &
1985 64714 : lone_keyword_l_val=.TRUE.)
1986 64714 : CALL section_add_keyword(section, keyword)
1987 64714 : CALL keyword_release(keyword)
1988 :
1989 : CALL keyword_create( &
1990 : keyword, __LOCATION__, &
1991 : name="K_MESH_G_FACTOR", &
1992 : description="The k-mesh for the Green's function can be chosen to be larger than the k-mesh for "// &
1993 : "W (without much higher computational cost). The factor given here multiplies the mesh for W to obtain "// &
1994 : "the k-mesh for G. Example: factor 4, k-mesh for W: 4x4x1 -> k-mesh for G: 16x16x1 (z-dir. is "// &
1995 : "non-periodic).", &
1996 64714 : default_i_val=1)
1997 64714 : CALL section_add_keyword(section, keyword)
1998 64714 : CALL keyword_release(keyword)
1999 :
2000 : CALL keyword_create( &
2001 : keyword, __LOCATION__, &
2002 : name="MIN_BLOCK_SIZE", &
2003 : description="Minimum tensor block size. Adjusting this value may have minor effect on "// &
2004 : "performance but default should be good enough.", &
2005 64714 : default_i_val=5)
2006 64714 : CALL section_add_keyword(section, keyword)
2007 64714 : CALL keyword_release(keyword)
2008 :
2009 : CALL keyword_create( &
2010 : keyword, __LOCATION__, &
2011 : name="MIN_BLOCK_SIZE_MO", &
2012 : description="Tensor block size for MOs. Only relevant for GW calculations. "// &
2013 : "The memory consumption of GW scales as O(MIN_BLOCK_SIZE_MO). It is recommended to "// &
2014 : "set this parameter to a smaller number if GW runs out of memory. "// &
2015 : "Otherwise the default should not be changed.", &
2016 64714 : default_i_val=64)
2017 64714 : CALL section_add_keyword(section, keyword)
2018 64714 : CALL keyword_release(keyword)
2019 :
2020 64714 : NULLIFY (subsection)
2021 64714 : CALL create_low_scaling_cphf(subsection)
2022 64714 : CALL section_add_subsection(section, subsection)
2023 64714 : CALL section_release(subsection)
2024 :
2025 64714 : END SUBROUTINE
2026 :
2027 : ! **************************************************************************************************
2028 : !> \brief ...
2029 : !> \param section ...
2030 : ! **************************************************************************************************
2031 64714 : SUBROUTINE create_wfc_gpw(section)
2032 : TYPE(section_type), POINTER :: section
2033 :
2034 : TYPE(keyword_type), POINTER :: keyword
2035 :
2036 64714 : CPASSERT(.NOT. ASSOCIATED(section))
2037 : CALL section_create(section, __LOCATION__, name="WFC_GPW", &
2038 : description="Parameters for the GPW approach in Wavefunction-based Correlation methods", &
2039 64714 : n_keywords=5, n_subsections=0, repeats=.FALSE.)
2040 :
2041 64714 : NULLIFY (keyword)
2042 : CALL keyword_create(keyword, __LOCATION__, name="EPS_GRID", &
2043 : description="Determines a threshold for the GPW based integration", &
2044 : usage="EPS_GRID 1.0E-9 ", type_of_var=real_t, &
2045 64714 : default_r_val=1.0E-8_dp)
2046 64714 : CALL section_add_keyword(section, keyword)
2047 64714 : CALL keyword_release(keyword)
2048 :
2049 : CALL keyword_create( &
2050 : keyword, __LOCATION__, name="EPS_FILTER", &
2051 : description="Determines a threshold for the DBCSR based multiply (usually 10 times smaller than EPS_GRID). "// &
2052 : "Normally, this EPS_FILTER determines accuracy and timing of cubic-scaling RPA calculation.", &
2053 : usage="EPS_FILTER 1.0E-10 ", type_of_var=real_t, &
2054 64714 : default_r_val=1.0E-9_dp)
2055 64714 : CALL section_add_keyword(section, keyword)
2056 64714 : CALL keyword_release(keyword)
2057 :
2058 : CALL keyword_create(keyword, __LOCATION__, name="CUTOFF", &
2059 : description="The cutoff of the finest grid level in the MP2 gpw integration.", &
2060 : usage="CUTOFF 300", type_of_var=real_t, &
2061 64714 : default_r_val=300.0_dp)
2062 64714 : CALL section_add_keyword(section, keyword)
2063 64714 : CALL keyword_release(keyword)
2064 :
2065 : CALL keyword_create(keyword, __LOCATION__, name="REL_CUTOFF", &
2066 : variants=(/"RELATIVE_CUTOFF"/), &
2067 : description="Determines the grid at which a Gaussian is mapped.", &
2068 : usage="REL_CUTOFF 50", type_of_var=real_t, &
2069 129428 : default_r_val=50.0_dp)
2070 64714 : CALL section_add_keyword(section, keyword)
2071 64714 : CALL keyword_release(keyword)
2072 :
2073 : CALL keyword_create(keyword, __LOCATION__, name="PRINT_LEVEL", &
2074 : variants=(/"IOLEVEL"/), &
2075 : description="How much output is written by the individual groups.", &
2076 : usage="PRINT_LEVEL HIGH", &
2077 : default_i_val=silent_print_level, enum_c_vals= &
2078 : s2a("SILENT", "LOW", "MEDIUM", "HIGH", "DEBUG"), &
2079 : enum_desc=s2a("Almost no output", &
2080 : "Little output", "Quite some output", "Lots of output", &
2081 : "Everything is written out, useful for debugging purposes only"), &
2082 : enum_i_vals=(/silent_print_level, low_print_level, medium_print_level, &
2083 129428 : high_print_level, debug_print_level/))
2084 64714 : CALL section_add_keyword(section, keyword)
2085 64714 : CALL keyword_release(keyword)
2086 :
2087 : CALL keyword_create( &
2088 : keyword, __LOCATION__, name="EPS_PGF_ORB_S", &
2089 : description="Screening for overlap matrix in RI. Usually, it is best to choose this parameter "// &
2090 : "to be very small since the inversion of overlap matrix might be ill-conditioned.", &
2091 : usage="EPS_PGF_ORB_S 1.0E-10 ", type_of_var=real_t, &
2092 64714 : default_r_val=1.0E-10_dp)
2093 64714 : CALL section_add_keyword(section, keyword)
2094 64714 : CALL keyword_release(keyword)
2095 :
2096 64714 : END SUBROUTINE create_wfc_gpw
2097 :
2098 : ! **************************************************************************************************
2099 : !> \brief ...
2100 : !> \param section ...
2101 : ! **************************************************************************************************
2102 64714 : SUBROUTINE create_cphf(section)
2103 : TYPE(section_type), POINTER :: section
2104 :
2105 : TYPE(keyword_type), POINTER :: keyword
2106 :
2107 64714 : CPASSERT(.NOT. ASSOCIATED(section))
2108 : CALL section_create( &
2109 : section, __LOCATION__, name="CPHF", &
2110 : description="Parameters influencing the solution of the Z-vector equations in MP2 gradients calculations.", &
2111 : n_keywords=2, n_subsections=0, repeats=.FALSE., &
2112 129428 : citations=(/DelBen2013/))
2113 :
2114 64714 : NULLIFY (keyword)
2115 :
2116 : CALL keyword_create(keyword, __LOCATION__, name="MAX_ITER", &
2117 : variants=(/"MAX_NUM_ITER"/), &
2118 : description="Maximum number of iterations allowed for the solution of the Z-vector equations.", &
2119 : usage="MAX_ITER 50", &
2120 129428 : default_i_val=30)
2121 64714 : CALL section_add_keyword(section, keyword)
2122 64714 : CALL keyword_release(keyword)
2123 :
2124 : CALL keyword_create(keyword, __LOCATION__, name="RESTART_EVERY", &
2125 : description="Restart iteration every given number of steps.", &
2126 : usage="RESTART_EVERY 5", &
2127 64714 : default_i_val=5)
2128 64714 : CALL section_add_keyword(section, keyword)
2129 64714 : CALL keyword_release(keyword)
2130 :
2131 : CALL keyword_create(keyword, __LOCATION__, name="SOLVER_METHOD", &
2132 : description="Chose solver of the z-vector equations.", &
2133 : usage="SOLVER_METHOD POPLE", enum_c_vals= &
2134 : s2a("POPLE", "CG", "RICHARDSON", "SD"), &
2135 : enum_desc=s2a("Pople's method (Default).", &
2136 : "Conjugated gradient method (equivalent to Pople).", &
2137 : "Richardson iteration", &
2138 : "Steepest Descent iteration"), &
2139 : enum_i_vals=(/z_solver_pople, z_solver_cg, z_solver_richardson, z_solver_sd/), &
2140 64714 : default_i_val=z_solver_pople)
2141 64714 : CALL section_add_keyword(section, keyword)
2142 64714 : CALL keyword_release(keyword)
2143 :
2144 : CALL keyword_create(keyword, __LOCATION__, name="EPS_CONV", &
2145 : description="Convergence threshold for the solution of the Z-vector equations. "// &
2146 : "The Z-vector equations have the form of a linear system of equations Ax=b, "// &
2147 : "convergence is achieved when |Ax-b|<=EPS_CONV.", &
2148 : usage="EPS_CONV 1.0E-6", type_of_var=real_t, &
2149 64714 : default_r_val=1.0E-4_dp)
2150 64714 : CALL section_add_keyword(section, keyword)
2151 64714 : CALL keyword_release(keyword)
2152 :
2153 : CALL keyword_create(keyword, __LOCATION__, name="SCALE_STEP_SIZE", &
2154 : description="Scaling factor of each step.", &
2155 : usage="SCALE_STEP_SIZE 1.0", &
2156 64714 : default_r_val=1.0_dp)
2157 64714 : CALL section_add_keyword(section, keyword)
2158 64714 : CALL keyword_release(keyword)
2159 :
2160 : CALL keyword_create(keyword, __LOCATION__, name="ENFORCE_DECREASE", &
2161 : description="Restarts if residual does not decrease.", &
2162 : usage="ENFORCE_DECREASE T", &
2163 : lone_keyword_l_val=.TRUE., &
2164 64714 : default_l_val=.FALSE.)
2165 64714 : CALL section_add_keyword(section, keyword)
2166 64714 : CALL keyword_release(keyword)
2167 :
2168 : CALL keyword_create(keyword, __LOCATION__, name="DO_POLAK_RIBIERE", &
2169 : description="Use a Polak-Ribiere update of the search vector in CG instead of the Fletcher "// &
2170 : "Reeves update. Improves the convergence with modified step sizes. "// &
2171 : "Ignored with other methods than CG.", &
2172 : usage="DO_POLAK_RIBIERE T", &
2173 : lone_keyword_l_val=.TRUE., &
2174 64714 : default_l_val=.FALSE.)
2175 64714 : CALL section_add_keyword(section, keyword)
2176 64714 : CALL keyword_release(keyword)
2177 :
2178 : CALL keyword_create(keyword, __LOCATION__, name="RECALC_RESIDUAL", &
2179 : description="Recalculates residual in every step.", &
2180 : usage="RECALC_RESIDUAL T", &
2181 : lone_keyword_l_val=.TRUE., &
2182 64714 : default_l_val=.FALSE.)
2183 64714 : CALL section_add_keyword(section, keyword)
2184 64714 : CALL keyword_release(keyword)
2185 :
2186 64714 : END SUBROUTINE create_cphf
2187 :
2188 : ! **************************************************************************************************
2189 : !> \brief ...
2190 : !> \param section ...
2191 : ! **************************************************************************************************
2192 64714 : SUBROUTINE create_low_scaling_cphf(section)
2193 : TYPE(section_type), POINTER :: section
2194 :
2195 : TYPE(keyword_type), POINTER :: keyword
2196 :
2197 64714 : NULLIFY (keyword)
2198 :
2199 64714 : CPASSERT(.NOT. ASSOCIATED(section))
2200 : CALL section_create(section, __LOCATION__, name="CPHF", &
2201 : description="Parameters influencing the solution of the Z-vector equations "// &
2202 : "in low-scaling Laplace-SOS-MP2 gradients calculations.", &
2203 64714 : n_keywords=5, n_subsections=0, repeats=.FALSE.)
2204 :
2205 : CALL keyword_create(keyword, __LOCATION__, name="EPS_CONV", &
2206 : description="Target accuracy for Z-vector euation solution.", &
2207 64714 : usage="EPS_CONV 1.e-6", default_r_val=1.e-6_dp)
2208 64714 : CALL section_add_keyword(section, keyword)
2209 64714 : CALL keyword_release(keyword)
2210 :
2211 : CALL keyword_create(keyword, __LOCATION__, name="MAX_ITER", &
2212 : description="Maximum number of conjugate gradient iteration to be performed for one optimization.", &
2213 64714 : usage="MAX_ITER 200", default_i_val=50)
2214 64714 : CALL section_add_keyword(section, keyword)
2215 64714 : CALL keyword_release(keyword)
2216 :
2217 : CALL keyword_create( &
2218 : keyword, __LOCATION__, name="PRECONDITIONER", &
2219 : description="Type of preconditioner to be used with all minimization schemes. "// &
2220 : "They differ in effectiveness, cost of construction, cost of application. "// &
2221 : "Properly preconditioned minimization can be orders of magnitude faster than doing nothing.", &
2222 : usage="PRECONDITIONER FULL_ALL", &
2223 : default_i_val=ot_precond_full_all, &
2224 : enum_c_vals=s2a("FULL_ALL", "FULL_SINGLE_INVERSE", "FULL_SINGLE", "FULL_KINETIC", "FULL_S_INVERSE", &
2225 : "NONE"), &
2226 : enum_desc=s2a("Most effective state selective preconditioner based on diagonalization, "// &
2227 : "requires the ENERGY_GAP parameter to be an underestimate of the HOMO-LUMO gap. "// &
2228 : "This preconditioner is recommended for almost all systems, except very large systems where "// &
2229 : "make_preconditioner would dominate the total computational cost.", &
2230 : "Based on H-eS cholesky inversion, similar to FULL_SINGLE in preconditioning efficiency "// &
2231 : "but cheaper to construct, "// &
2232 : "might be somewhat less robust. Recommended for large systems.", &
2233 : "Based on H-eS diagonalisation, not as good as FULL_ALL, but somewhat cheaper to apply. ", &
2234 : "Cholesky inversion of S and T, fast construction, robust, and relatively good, "// &
2235 : "use for very large systems.", &
2236 : "Cholesky inversion of S, not as good as FULL_KINETIC, yet equally expensive.", &
2237 : "skip preconditioning"), &
2238 : enum_i_vals=(/ot_precond_full_all, ot_precond_full_single_inverse, ot_precond_full_single, &
2239 64714 : ot_precond_full_kinetic, ot_precond_s_inverse, ot_precond_none/))
2240 64714 : CALL section_add_keyword(section, keyword)
2241 64714 : CALL keyword_release(keyword)
2242 :
2243 : CALL keyword_create(keyword, __LOCATION__, name="ENERGY_GAP", &
2244 : description="Energy gap estimate [a.u.] for preconditioning", &
2245 : usage="ENERGY_GAP 0.1", &
2246 64714 : default_r_val=0.2_dp)
2247 64714 : CALL section_add_keyword(section, keyword)
2248 64714 : CALL keyword_release(keyword)
2249 :
2250 64714 : END SUBROUTINE create_low_scaling_cphf
2251 :
2252 : ! **************************************************************************************************
2253 : !> \brief ...
2254 : !> \param section ...
2255 : ! **************************************************************************************************
2256 64714 : SUBROUTINE create_mp2_potential(section)
2257 : TYPE(section_type), POINTER :: section
2258 :
2259 : TYPE(keyword_type), POINTER :: keyword
2260 :
2261 64714 : CPASSERT(.NOT. ASSOCIATED(section))
2262 : CALL section_create(section, __LOCATION__, name="INTERACTION_POTENTIAL", &
2263 : description="Parameters the interaction potential in computing the biel integrals", &
2264 64714 : n_keywords=4, n_subsections=0, repeats=.FALSE.)
2265 :
2266 64714 : NULLIFY (keyword)
2267 : CALL keyword_create( &
2268 : keyword, __LOCATION__, &
2269 : name="POTENTIAL_TYPE", &
2270 : description="Which interaction potential should be used "// &
2271 : "(Coulomb, TShPSC operator).", &
2272 : usage="POTENTIAL_TYPE TSHPSC", &
2273 : enum_c_vals=s2a("COULOMB", "TShPSC", "LONGRANGE", "SHORTRANGE", "TRUNCATED", "MIX_CL", "IDENTITY"), &
2274 : enum_i_vals=(/do_potential_coulomb, &
2275 : do_potential_TShPSC, &
2276 : do_potential_long, &
2277 : do_potential_short, &
2278 : do_potential_truncated, &
2279 : do_potential_mix_cl, &
2280 : do_potential_id/), &
2281 : enum_desc=s2a("Coulomb potential: 1/r", &
2282 : "| Range | TShPSC |"//newline// &
2283 : "| ----- | ------ |"//newline// &
2284 : "| $ x \leq R_c $ | $ 1/x - s/R_c $ |"//newline// &
2285 : "| $ R_c < x \leq nR_c $ | "// &
2286 : "$ (1 - s)/R_c - (x - R_c)/R_c^2 + (x - R_c)^2/R_c^3 - "// &
2287 : "(2n^2 - 7n + 9 - 4s)(x - R_c)^3/(R_c^4(n^2 - 2n + 1)(n - 1)) + "// &
2288 : "(6-3s - 4n + n^2)(x - R_c)^4/(R_c^5(n^4 - 4n^3 + 6n^2 - 4n + 1)) $ "// &
2289 : "(4th order polynomial) | "//newline// &
2290 : "| $ x > nR_c $ | $ 0 $ | "//newline, &
2291 : "Longrange Coulomb potential: $ \operatorname{erf}(wr)/r $", &
2292 : "Shortrange Coulomb potential: $ \operatorname{erfc}(wr)/r $", &
2293 : "Truncated Coulomb potential", &
2294 : "Mixed Coulomb/Longrange Coulomb potential", &
2295 : "Delta potential"), &
2296 64714 : default_i_val=do_potential_coulomb)
2297 64714 : CALL section_add_keyword(section, keyword)
2298 64714 : CALL keyword_release(keyword)
2299 :
2300 : CALL keyword_create(keyword, __LOCATION__, name="TRUNCATION_RADIUS", &
2301 : variants=(/"CUTOFF_RADIUS"/), &
2302 : description="Determines truncation radius for the truncated potentials. "// &
2303 : "Only valid when doing truncated calculations", &
2304 : usage="TRUNCATION_RADIUS 10.0", type_of_var=real_t, &
2305 : default_r_val=10.0_dp, &
2306 129428 : unit_str="angstrom")
2307 64714 : CALL section_add_keyword(section, keyword)
2308 64714 : CALL keyword_release(keyword)
2309 :
2310 : CALL keyword_create( &
2311 : keyword, __LOCATION__, &
2312 : name="POTENTIAL_DATA", &
2313 : variants=s2a("TShPSC_DATA", "T_C_G_DATA"), &
2314 : description="Location of the file TShPSC.dat or t_c_g.dat that contains the data for the "// &
2315 : "evaluation of the evaluation of the truncated potentials", &
2316 : usage="TShPSC_DATA t_sh_p_s_c.dat", &
2317 64714 : default_c_val="t_sh_p_s_c.dat")
2318 64714 : CALL section_add_keyword(section, keyword)
2319 64714 : CALL keyword_release(keyword)
2320 :
2321 : CALL keyword_create( &
2322 : keyword, __LOCATION__, &
2323 : name="OMEGA", &
2324 : description="Range separation parameter for the longrange or shortrange potential. "// &
2325 : "Only valid when longrange or shortrange potential is requested.", &
2326 : usage="OMEGA 0.5", type_of_var=real_t, &
2327 64714 : default_r_val=0.5_dp)
2328 64714 : CALL section_add_keyword(section, keyword)
2329 64714 : CALL keyword_release(keyword)
2330 :
2331 : CALL keyword_create( &
2332 : keyword, __LOCATION__, &
2333 : name="SCALE_COULOMB", &
2334 : description="Scaling factor of (truncated) Coulomb potential in mixed (truncated) Coulomb/Longrange potential. "// &
2335 : "Only valid when mixed potential is requested.", &
2336 : usage="SCALE_COULOMB 0.5", type_of_var=real_t, &
2337 64714 : default_r_val=1.0_dp)
2338 64714 : CALL section_add_keyword(section, keyword)
2339 64714 : CALL keyword_release(keyword)
2340 :
2341 : CALL keyword_create( &
2342 : keyword, __LOCATION__, &
2343 : name="SCALE_LONGRANGE", &
2344 : description="Scaling factor of longrange Coulomb potential in mixed (truncated) Coulomb/Longrange potential. "// &
2345 : "Only valid when mixed potential is requested.", &
2346 : usage="SCALE_LONGRANGE 0.5", type_of_var=real_t, &
2347 64714 : default_r_val=1.0_dp)
2348 64714 : CALL section_add_keyword(section, keyword)
2349 64714 : CALL keyword_release(keyword)
2350 :
2351 64714 : END SUBROUTINE create_mp2_potential
2352 :
2353 : ! **************************************************************************************************
2354 : !> \brief ...
2355 : !> \param section ...
2356 : ! **************************************************************************************************
2357 64714 : SUBROUTINE create_ri_section(section)
2358 : TYPE(section_type), POINTER :: section
2359 :
2360 : TYPE(keyword_type), POINTER :: keyword
2361 : TYPE(section_type), POINTER :: subsection
2362 :
2363 64714 : CPASSERT(.NOT. ASSOCIATED(section))
2364 : CALL section_create(section, __LOCATION__, name="RI", &
2365 : description="Parameters influencing resolution of the identity (RI) that is "// &
2366 : "used in RI-MP2, RI-RPA, RI-SOS-MP2 and GW (inside RI-RPA).", &
2367 64714 : n_keywords=6, n_subsections=2, repeats=.FALSE.)
2368 :
2369 64714 : NULLIFY (subsection)
2370 64714 : CALL create_RI_metric_section(subsection)
2371 64714 : CALL section_add_subsection(section, subsection)
2372 64714 : CALL section_release(subsection)
2373 :
2374 64714 : CALL create_opt_ri_basis(subsection)
2375 64714 : CALL section_add_subsection(section, subsection)
2376 64714 : CALL section_release(subsection)
2377 :
2378 64714 : NULLIFY (keyword)
2379 : CALL keyword_create( &
2380 : keyword, __LOCATION__, &
2381 : name="ROW_BLOCK", &
2382 : variants=(/"ROW_BLOCK_SIZE"/), &
2383 : description="Size of the row block used in the SCALAPACK block cyclic data distribution. "// &
2384 : "Default is (ROW_BLOCK=-1) is automatic. "// &
2385 : "A proper choice can speedup the parallel matrix multiplication in the case of RI-RPA and RI-SOS-MP2-Laplace.", &
2386 : usage="ROW_BLOCK 512", &
2387 129428 : default_i_val=-1)
2388 64714 : CALL section_add_keyword(section, keyword)
2389 64714 : CALL keyword_release(keyword)
2390 :
2391 : CALL keyword_create( &
2392 : keyword, __LOCATION__, &
2393 : name="COL_BLOCK", &
2394 : variants=(/"COL_BLOCK_SIZE"/), &
2395 : description="Size of the column block used in the SCALAPACK block cyclic data distribution. "// &
2396 : "Default is (COL_BLOCK=-1) is automatic. "// &
2397 : "A proper choice can speedup the parallel matrix multiplication in the case of RI-RPA and RI-SOS-MP2-Laplace.", &
2398 : usage="COL_BLOCK 512", &
2399 129428 : default_i_val=-1)
2400 64714 : CALL section_add_keyword(section, keyword)
2401 64714 : CALL keyword_release(keyword)
2402 :
2403 : CALL keyword_create( &
2404 : keyword, __LOCATION__, &
2405 : name="CALC_COND_NUM", &
2406 : variants=(/"CALC_CONDITION_NUMBER"/), &
2407 : description="Calculate the condition number of the (P|Q) matrix for the RI methods.", &
2408 : usage="CALC_COND_NUM", &
2409 : default_l_val=.FALSE., &
2410 129428 : lone_keyword_l_val=.TRUE.)
2411 64714 : CALL section_add_keyword(section, keyword)
2412 64714 : CALL keyword_release(keyword)
2413 :
2414 : CALL keyword_create(keyword, __LOCATION__, name="DO_SVD", &
2415 : description="Wether to perform a singular value decomposition instead of the Cholesky decomposition "// &
2416 : "of the potential operator in the RI basis. Computationally expensive but numerically more stable. "// &
2417 : "It reduces the computational costs of some subsequent steps. Recommended when a longrange Coulomb "// &
2418 : "potential is employed.", &
2419 : usage="DO_SVD .TRUE.", &
2420 64714 : default_l_val=.FALSE., lone_keyword_l_val=.TRUE.)
2421 64714 : CALL section_add_keyword(section, keyword)
2422 64714 : CALL keyword_release(keyword)
2423 :
2424 : CALL keyword_create(keyword, __LOCATION__, name="EPS_SVD", &
2425 : description="Determines the upper bound of eigenvectors to be removed during the SVD (see DO_SVD).", &
2426 : usage="EPS_SVD 1E-5", &
2427 64714 : default_r_val=0.0_dp)
2428 64714 : CALL section_add_keyword(section, keyword)
2429 64714 : CALL keyword_release(keyword)
2430 :
2431 : CALL keyword_create(keyword, __LOCATION__, name="ERI_BLKSIZE", &
2432 : description="block sizes for tensors (only used if ERI_METHOD=MME). First value "// &
2433 : "is the block size for ORB basis, second value is the block size for RI_AUX basis.", &
2434 : usage="ERI_BLKSIZE", &
2435 : n_var=2, &
2436 64714 : default_i_vals=(/4, 16/))
2437 64714 : CALL section_add_keyword(section, keyword)
2438 64714 : CALL keyword_release(keyword)
2439 :
2440 64714 : END SUBROUTINE create_ri_section
2441 :
2442 : ! **************************************************************************************************
2443 : !> \brief ...
2444 : !> \param section ...
2445 : ! **************************************************************************************************
2446 64714 : SUBROUTINE create_integrals_section(section)
2447 : TYPE(section_type), POINTER :: section
2448 :
2449 : TYPE(keyword_type), POINTER :: keyword
2450 : TYPE(section_type), POINTER :: subsection
2451 :
2452 64714 : CPASSERT(.NOT. ASSOCIATED(section))
2453 : CALL section_create(section, __LOCATION__, name="INTEGRALS", &
2454 : description="Parameters controlling how to compute integrals that are needed "// &
2455 : "in MP2, RI-MP2, RI-RPA, RI-SOS-MP2 and GW (inside RI-RPA).", &
2456 64714 : n_keywords=2, n_subsections=3, repeats=.FALSE.)
2457 :
2458 64714 : NULLIFY (subsection)
2459 64714 : CALL create_eri_mme_section(subsection)
2460 64714 : CALL section_add_subsection(section, subsection)
2461 64714 : CALL section_release(subsection)
2462 :
2463 64714 : CALL create_wfc_gpw(subsection)
2464 64714 : CALL section_add_subsection(section, subsection)
2465 64714 : CALL section_release(subsection)
2466 :
2467 64714 : CALL create_mp2_potential(subsection)
2468 64714 : CALL section_add_subsection(section, subsection)
2469 64714 : CALL section_release(subsection)
2470 :
2471 64714 : NULLIFY (keyword)
2472 : CALL keyword_create(keyword, __LOCATION__, name="ERI_METHOD", &
2473 : description="Method for calculating periodic electron repulsion integrals "// &
2474 : "(MME method is faster but experimental, forces not yet implemented). "// &
2475 : "Obara-Saika (OS) for the Coulomb operator can only be used for non-periodic calculations.", &
2476 : usage="ERI_METHOD MME", &
2477 : enum_c_vals=s2a("DEFAULT", "GPW", "MME", "OS"), &
2478 : enum_i_vals=(/eri_default, do_eri_gpw, do_eri_mme, do_eri_os/), &
2479 : enum_desc=s2a("Use default ERI method (for periodic systems: GPW, for molecules: OS, "// &
2480 : "for MP2 and RI-MP2: GPW in any case).", &
2481 : "Uses Gaussian Plane Wave method [DelBen2013].", &
2482 : "Uses MiniMax-Ewald method (experimental, ERI_MME subsection, only for fully periodic "// &
2483 : "systems with orthorhombic cells).", &
2484 : "Use analytical Obara-Saika method."), &
2485 64714 : default_i_val=eri_default)
2486 64714 : CALL section_add_keyword(section, keyword)
2487 64714 : CALL keyword_release(keyword)
2488 :
2489 : CALL keyword_create(keyword, __LOCATION__, name="SIZE_LATTICE_SUM", &
2490 : description="Size of sum range L. ", &
2491 : usage="SIZE_LATTICE_SUM 10", &
2492 64714 : default_i_val=5)
2493 64714 : CALL section_add_keyword(section, keyword)
2494 64714 : CALL keyword_release(keyword)
2495 :
2496 64714 : END SUBROUTINE create_integrals_section
2497 :
2498 : ! **************************************************************************************************
2499 : !> \brief ...
2500 : !> \param section ...
2501 : ! **************************************************************************************************
2502 64714 : SUBROUTINE create_RI_metric_section(section)
2503 : TYPE(section_type), POINTER :: section
2504 :
2505 : TYPE(keyword_type), POINTER :: keyword
2506 :
2507 64714 : CPASSERT(.NOT. ASSOCIATED(section))
2508 : CALL section_create(section, __LOCATION__, name="RI_METRIC", &
2509 : description="Sets up RI metric", &
2510 64714 : repeats=.FALSE.)
2511 :
2512 64714 : NULLIFY (keyword)
2513 : CALL keyword_create( &
2514 : keyword, __LOCATION__, &
2515 : name="POTENTIAL_TYPE", &
2516 : description="Decides which operator/metric is used for resolution of the identity (RI).", &
2517 : usage="POTENTIAL_TYPE DEFAULT", &
2518 : enum_c_vals=s2a("DEFAULT", "COULOMB", "IDENTITY", "LONGRANGE", "SHORTRANGE", "TRUNCATED"), &
2519 : enum_i_vals=(/ri_default, do_potential_coulomb, do_potential_id, do_potential_long, &
2520 : do_potential_short, do_potential_truncated/), &
2521 : enum_desc=s2a("Use Coulomb metric for RI-MP2 and normal-scaling RI-SOS-MP2, RI-RPA and GW. "// &
2522 : "Use Overlap metric for low-scaling RI-SOS-MP2, RI-RPA and GW for periodic systems. "// &
2523 : "Use truncated Coulomb metric for low-scaling RI-SOS-MP2, RI-RPA and GW for non-periodic systems.", &
2524 : "Coulomb metric: 1/r. Recommended for RI-MP2,", &
2525 : "Overlap metric: delta(r).", &
2526 : "Longrange metric: erf(omega*r)/r. Not recommended with DO_SVD .TRUE.", &
2527 : "Shortrange metric: erfc(omega*r)/r", &
2528 : "Truncated Coulomb metric: if (r < R_c) 1/r else 0. More "// &
2529 : "accurate than IDENTITY for non-periodic systems. Recommended for low-scaling methods."), &
2530 64714 : default_i_val=ri_default)
2531 64714 : CALL section_add_keyword(section, keyword)
2532 64714 : CALL keyword_release(keyword)
2533 :
2534 64714 : NULLIFY (keyword)
2535 : CALL keyword_create( &
2536 : keyword, __LOCATION__, &
2537 : name="OMEGA", &
2538 : description="The range parameter for the short/long range operator (in 1/a0).", &
2539 : usage="OMEGA 0.5", &
2540 64714 : default_r_val=0.0_dp)
2541 64714 : CALL section_add_keyword(section, keyword)
2542 64714 : CALL keyword_release(keyword)
2543 :
2544 : CALL keyword_create(keyword, __LOCATION__, name="CUTOFF_RADIUS", &
2545 : description="The cutoff radius (in Angstrom) for the truncated Coulomb operator.", &
2546 : usage="CUTOFF_RADIUS 3.0", default_r_val=cp_unit_to_cp2k(value=3.0_dp, unit_str="angstrom"), &
2547 64714 : type_of_var=real_t, unit_str="angstrom")
2548 64714 : CALL section_add_keyword(section, keyword)
2549 64714 : CALL keyword_release(keyword)
2550 :
2551 : CALL keyword_create( &
2552 : keyword, __LOCATION__, &
2553 : name="T_C_G_DATA", &
2554 : description="Location of the file t_c_g.dat that contains the data for the "// &
2555 : "evaluation of the truncated gamma function ", &
2556 64714 : default_c_val="t_c_g.dat")
2557 64714 : CALL section_add_keyword(section, keyword)
2558 64714 : CALL keyword_release(keyword)
2559 :
2560 : CALL keyword_create(keyword, __LOCATION__, name="EPS_RANGE", &
2561 : description="The threshold to determine the effective range of the short range "// &
2562 : "RI metric: erfc(omega*eff_range)/eff_range = EPS_RANGE", &
2563 : default_r_val=1.0E-08_dp, &
2564 64714 : repeats=.FALSE.)
2565 64714 : CALL section_add_keyword(section, keyword)
2566 64714 : CALL keyword_release(keyword)
2567 :
2568 64714 : END SUBROUTINE create_RI_metric_section
2569 :
2570 : END MODULE input_cp2k_mp2
|
