mu_initial_guess

[KeyuAN95]

Dear Alex and all,

I run the solid_dmft with the following dmft_config.toml:
#######################################
[general]
seedname = "test_wannier"
jobname = "U3.1"
beta = 40
n_iw = 251
n_tau = 5001
n_iter_dmft = 10
prec_mu = 0.0001
mu_initial_guess = 13.3115
h_int_type = "kanamori"
U = 3.1
J = 0.0

dc_type = 0
dc_dmft = false
h5_save_freq = 1
[solver]
type = "cthyb"
length_cycle = 60
n_warmup_cycles = 1e+4
n_cycles_tot = 2e+6
imag_threshold = 1e-6
measure_density_matrix = true
delta_interface=true
perform_tail_fit = true
####################################
it | mu | G(beta/2) per orbital | orbital occs up+down | impurity occ
0 | 0.55267 | -0.04555 -0.06437 | 0.98401 0.97000 | 1.95402
1 | 0.55267 | -0.04519 -0.06516 | 1.00095 0.95128 | 1.95223
2 | 0.54320 | -0.04623 -0.06262 | 0.97786 0.95671 | 1.93457
3 | 0.54569 | -0.04363 -0.06560 | 0.96887 0.94197 | 1.91084
4 | 0.55155 | -0.04418 -0.06334 | 0.98258 0.95281 | 1.93539
5 | 0.56119 | -0.04419 -0.06583 | 0.96832 0.94570 | 1.91402
6 | 0.56700 | -0.04438 -0.06588 | 0.96992 0.95199 | 1.92191
7 | 0.57212 | -0.04386 -0.06572 | 0.99978 0.98294 | 1.98273
8 | 0.56013 | -0.04249 -0.06476 | 0.99534 0.94508 | 1.94041
9 | 0.56597 | -0.04259 -0.06269 | 0.97671 0.96611 | 1.94282
10 | 0.56296 | -0.04551 -0.06560 | 0.99362 0.97021 | 1.96383
11 | 0.55860 | -0.04308 -0.06911 | 0.97626 0.95802 | 1.93428
12 | 0.54908 | -0.04250 -0.06567 | 0.99067 0.96105 | 1.95172
13 | 0.54729 | -0.04468 -0.06503 | 0.98642 0.94993 | 1.93635
14 | 0.54440 | -0.04401 -0.06558 | 0.97823 0.95619 | 1.93442
15 | 0.54615 | -0.04377 -0.06541 | 0.98066 0.94751 | 1.92818
16 | 0.54934 | -0.04522 -0.06455 | 0.97793 0.95845 | 1.93638
17 | 0.55049 | -0.04417 -0.06407 | 0.97252 0.95499 | 1.92751
18 | 0.55487 | -0.04380 -0.06579 | 0.96830 0.96988 | 1.93818
19 | 0.55017 | -0.04273 -0.06703 | 0.96811 0.94844 | 1.91655
20 | 0.55858 | -0.04452 -0.06304 | 0.96412 0.94878 | 1.91291
21 | 0.55739 | -0.04497 -0.06566 | 0.98887 0.96901 | 1.95788
22 | 0.55346 | -0.04397 -0.06506 | 0.98979 0.95225 | 1.94204
23 | 0.55253 | -0.04305 -0.06582 | 0.97203 0.94926 | 1.92129
24 | 0.55842 | -0.04342 -0.06513 | 0.97468 0.96106 | 1.93574
25 | 0.56121 | -0.04355 -0.06468 | 0.98758 0.94158 | 1.92915
26 | 0.56668 | -0.04331 -0.06546 | 0.98298 0.94715 | 1.93013
27 | 0.56503 | -0.04498 -0.06393 | 0.98962 0.93672 | 1.92634
28 | 0.56664 | -0.04342 -0.06412 | 1.00418 0.95557 | 1.95975
29 | 0.56685 | -0.04350 -0.06405 | 0.98903 0.94795 | 1.93698
30 | 0.56777 | -0.04442 -0.06585 | 0.98341 0.96469 | 1.94809
##################################################################################
After the running, the chemical potential μ did not converge to a fixed value. How can I deal with this?

[KeyuAN95]

I can achieve convergence of the chemical potential mu to a fixed value by removing the Fermi energy in .inp.

[the-hampel]

Hi @KeyuAN95 ,

sorry for the slow response. Somehow I missed your post. Yes probably the chemical potential was not found correctly, since you specified an initial guess (13.xx) and the actual fermi level was now around 0.0. As you correctly noticed this is because adding it in the .inp file will subtract it when converting to the h5 format for solid_dmft.

Best,
Alex