The value of the population is very small?

[AtsushiTanimoto]

Thank you for maintaining this nice code. To calculate the N04 population data, we have performed the following code. Although we could obtain the N04a_t00d0i2.sp file, the value of the population in the ground state was very small (~1.0e-5). How can we obtain realistic values for the population of the ground state?

import numpy
import pfac.fac
import pfac.spm
import subprocess


if __name__=="__main__":
    atomic_number   = 7
    electron_number = 4
    densities       = 1e-10*numpy.logspace(0, 0, 1)
    temperatures    = 1e+00*numpy.logspace(1, 1, 1)
    atomic_symbol   = pfac.fac.ATOMICSYMBOL[atomic_number]
    input_dir       = "database01/{0:s}/".format(atomic_symbol)
    output_dir      = "database01/{0:s}/{0:s}{1:02d}_spec/".format(atomic_symbol, electron_number)
    populations     = len(temperatures)*[(1+atomic_number)*[1.0]]
    subprocess.run("rm -r {0:s}".format(output_dir), shell=True)
    subprocess.run("mkdir {0:s}".format(output_dir), shell=True)
    pfac.spm.spectrum(neles=[electron_number], temp=temperatures, den=densities, population=populations, pref=atomic_symbol, dir0=input_dir, dir1=output_dir, nion=2, ai=1, ce=1, ci=1, rr=1, rrc=1)

[mfgu]

I am not able to reproduce the problem described. please post the exact steps you use in the calculation.

[AtsushiTanimoto]

I am sorry for the lack of information. First, I calculated the basic atomic database of N03 and N04 by pfac.atom. I then computed the population data of N04 by pfac.spm. The Python code I ran is as follows. However, the obtained population value of the ground state in N04a_d00t00i2.sp is very small as shown below.

Environment:
Apple M1 Macbook Pro
macOS: Ventura 13.4.1
Python: Python 3.11.4

import numpy
import pfac.atom
import pfac.fac
import pfac.spm
import subprocess


if __name__=="__main__":
    atomic_number   = 7
    atomic_symbol   = pfac.fac.ATOMICSYMBOL[atomic_number]
    electron_number = 4
    densities       = 1e-10*numpy.logspace(1, 1, 1)
    temperatures    = 1e+00*numpy.logspace(1, 1, 1)
    populations     = len(temperatures)*[(1+atomic_number)*[1.0]]
    directory01     = "database/{0:s}/".format(atomic_symbol)
    directory02     = "database/{0:s}/".format(atomic_symbol)

    subprocess.run("mkdir {0:s}".format(directory01), shell=True)
    subprocess.run("mkdir {0:s}".format(directory02), shell=True)
    pfac.atom.atomic_data(nele=[electron_number-1, electron_number], asym=atomic_symbol, dir=directory01)
    pfac.spm.spectrum(neles=[electron_number], temp=temperatures, den=densities, population=populations, pref=atomic_symbol, dir0=directory01, dir1=directory02, nion=2, ai=1, ce=1, ci=1, rr=1, rrc=1)

FAC 1.1.5
Endian	= 0
TSess	= 1689041061
Type	= 7
Verbose	= 1
N Z	=   7.0
NBlocks	= 2319

NELE	= 4
NTRANS	= 1
TYPE	= 0000000
IBLK	= 6
ICOMP	= 1*2.2*2
FBLK	= 0
FCOMP	= 
     0      0  9.418330E+03  1.3418E-05  1.0000E+00  2.8015E-07

[mfgu]

i cannot reproduce the results with your script. the ground state population shows below in my run.
NELE = 4
NTRANS = 1
TYPE = 0000000
IBLK = 6
ICOMP = 12.22
FBLK = 0
FCOMP =
0 0 9.418330E+03 9.7816E-01 1.0000E+00 2.8015E-07

can you do two things.

1. try update FAC to the current version on github. if you still get the same result,
2. try to run the same script on a linux machine. maybe there is something peculiar about the apple M1/M2 chip, which I haven't had a chance to test.

[AtsushiTanimoto]

Thank you for your reply.

Try updating FAC to the current version on github. if you still get the same result,

Two weeks ago I was using the latest version of the flexible atomic code.
Today I updated the flexible atomic code. As a result, I get the following segmentation fault. How can I work around this?

ATOM = N, Z = 7
EN...
Structure: ground complex
Structure: ionized complexes
    ['ion.0.0']
    ['ion.1.0']
    ['ion.2.0']
Structure: excited complexes
    ['exc.0.4.0']
    ['exc.0.5.0']
    ['exc.0.6.0']
    ['exc.0.7.0']
    (['ion.0.0'], 8)
zsh: segmentation fault  /opt/homebrew/bin/python3 spectrum.py

[mfgu]

The seg.fault. was recently introduced. the new update has the fix.