fitjob_get_data

#!/usr/bin/env python3
# coding: utf-8

import sys
from pathlib import Path

try:
  import pandas as pd
except ModuleNotFoundError:
  sys.exit("Module 'pandas' not installed. Exiting.")

try:
  import numpy as np
except ModuleNotFoundError:
  sys.exit("Module 'numpy' not installed. Exiting.")


def main():
  # setting some variables
  omegas = [100, 150, 200, 250, 300, 350, 400, 450]
  num_jobs = len(omegas)
  cols = ["tmer", "dimer", "mono A", "monoB", "energy"]

  factor = 627.50947428
  num_decimals = 4
  out_file = f"res_{Path.cwd().name.lower()}"
  data = []

  # Reading in reference data
  nci = pd.read_csv("NCIBLIND10/ref.csv", dtype={"dimer": str})
  nci_list = nci["dimer"].to_list()
  nci_ref = nci["reference"].to_list()

  s22x5 = pd.read_csv("S22x5/ref.csv", dtype={"dimer": str})
  s22x5_list = s22x5["dimer"].to_list()
  s22x5_ref = s22x5["reference"].to_list()

  s66x8 = pd.read_csv("S66x8/ref.csv", dtype={"dimer": str})
  s66x8_list = s66x8["dimer"].to_list()
  s66x8_ref = s66x8["reference"].to_list()


  # S22x5
  for i, d in enumerate(s22x5_list):
    dimer = f"S22x5/{d}"
    monoA = f"S22x5/{d.split('_')[0]}A"
    monoB = f"S22x5/{d.split('_')[0]}B"

    e_dimer = []
    with open(Path(f"{dimer}/job.out"), errors='ignore') as f:
      count_criter = 0

      for line in f:
        if "Convergence criterion met" in line:
          energy = line.split()[1]
          e_dimer.append(float(energy))
          count_criter += 1

      # check convergence for my multijob
      if count_criter != num_jobs:
        sys.exit()

    e_monoA = []
    with open(Path(f"{monoA}/job.out"), errors='ignore') as f:
      count_criter = 0

      for line in f:
        if "Convergence criterion met" in line:
          energy = line.split()[1]
          e_monoA.append(float(energy))
          count_criter += 1

      # check convergence for my multijob
      if count_criter != num_jobs:
        sys.exit()

    e_monoB = []
    with open(Path(f"{monoB}/job.out"), errors='ignore') as f:
      count_criter = 0

      for line in f:
        if "Convergence criterion met" in line:
          energy = line.split()[1]
          e_monoB.append(float(energy))
          count_criter += 1

      # check convergence for my multijob
      if count_criter != num_jobs:
        sys.exit()

    temp = ["$tmer", f"{dimer}/${{f}}", f"{monoA}/${{f}}", f"{monoB}/${{f}}"]
    for j in range(num_jobs):
      e = (e_dimer[j] - e_monoA[j] - e_monoB[j])*factor - s22x5_ref[i] 
      temp.append(np.around(e, decimals=num_decimals))

    data.append(temp)


  # S66x8
  for i, d in enumerate(s66x8_list):
    dimer = f"S66x8/{d}"
    monoA = f"S66x8/{d.split('.')[0]}A"
    monoB = f"S66x8/{d.split('.')[0]}B"

    e_dimer = []
    with open(Path(f"{dimer}/job.out"), errors='ignore') as f:
      count_criter = 0

      for line in f:
        if "Convergence criterion met" in line:
          energy = line.split()[1]
          e_dimer.append(float(energy))
          count_criter += 1

      # check convergence for my multijob
      if count_criter != num_jobs:
        sys.exit()

    e_monoA = []
    with open(Path(f"{monoA}/job.out"), errors='ignore') as f:
      count_criter = 0

      for line in f:
        if "Convergence criterion met" in line:
          energy = line.split()[1]
          e_monoA.append(float(energy))
          count_criter += 1

      # check convergence for my multijob
      if count_criter != num_jobs:
        sys.exit()

    e_monoB = []
    with open(Path(f"{monoB}/job.out"), errors='ignore') as f:
      count_criter = 0

      for line in f:
        if "Convergence criterion met" in line:
          energy = line.split()[1]
          e_monoB.append(float(energy))
          count_criter += 1

      # check convergence for my multijob
      if count_criter != num_jobs:
        sys.exit()

    temp = ["$tmer", f"{dimer}/${{f}}", f"{monoA}/${{f}}", f"{monoB}/${{f}}"]
    for j in range(num_jobs):
      e = (e_dimer[j] - e_monoA[j] - e_monoB[j])*factor - s66x8_ref[i] 
      temp.append(np.around(e, decimals=num_decimals))

    data.append(temp)


  # NCIBLIND10
  for i, d in enumerate(nci_list):
    dimer = f"NCIBLIND10/{d}"

    if int(d) < 100:
      monoA = f"NCIBLIND10/{d[0]}A"
      monoB = f"NCIBLIND10/{d[0]}B"
    else:
      monoA = "NCIBLIND10/10A"
      monoB = "NCIBLIND10/10B"

    e_dimer = []
    with open(Path(f"{dimer}/job.out"), errors='ignore') as f:
      count_criter = 0

      for line in f:
        if "Convergence criterion met" in line:
          energy = line.split()[1]
          e_dimer.append(float(energy))
          count_criter += 1

      # check convergence for my multijob
      if count_criter != num_jobs:
        sys.exit()

    e_monoA = []
    with open(Path(f"{monoA}/job.out"), errors='ignore') as f:
      count_criter = 0

      for line in f:
        if "Convergence criterion met" in line:
          energy = line.split()[1]
          e_monoA.append(float(energy))
          count_criter += 1

      # check convergence for my multijob
      if count_criter != num_jobs:
        sys.exit()

    e_monoB = []
    with open(Path(f"{monoB}/job.out"), errors='ignore') as f:
      count_criter = 0

      for line in f:
        if "Convergence criterion met" in line:
          energy = line.split()[1]
          e_monoB.append(float(energy))
          count_criter += 1

      # check convergence for my multijob
      if count_criter != num_jobs:
        sys.exit()

    temp = ["$tmer", f"{dimer}/${{f}}", f"{monoA}/${{f}}", f"{monoB}/${{f}}"]
    for j in range(num_jobs):
      e = (e_dimer[j] - e_monoA[j] - e_monoB[j])*factor - nci_ref[i]
      temp.append(np.around(e, decimals=num_decimals))

    data.append(temp)


  # Save data
  data = np.asarray(data, dtype=str)
  for i, omega in enumerate(omegas):
    np.savetxt(f"{out_file}_{omega}.sh", data[:, [0, 1, 2, 3, i + 4]], delimiter=" ", fmt="%s")

    #df = pd.DataFrame(columns=cols, data=data[:, [0, 1, 2, 3, i + 4]])
    #df.to_csv(f"{out_file}_{omega}.sh", index=False, header=False, sep=" ")



if __name__ == '__main__':
  main()