formqc troubleshooting

troubleshoot1.py

#!/usr/bin/env python

from glob import glob
from os.path import isfile, basename
import pandas as pd

subjects=glob('/data/predict1/data_from_nda/Pronet/PHOENIX/PROTECTED/*/raw/*')
subjects=[basename(s) for s in subjects]

print('for how many subjects inclusionexclusion_criteria_review does not exist?')
count=0
for s in subjects:
    p=s[:2]+'-'+s[:9]+'-form_inclusionexclusion_criteria_review-day*to*csv'
    p1=glob(p)
    if not p1:
        count+=1

print(count)



print('for how many subjects sociodemographics does not exist')
count=0
for s in subjects:
    p=s[:2]+'-'+s[:9]+'-form_sociodemographics-day*to*csv'
    p1=glob(p)
    if not p1:
        count+=1

print(count)



print('for how many subjects either inclusionexclusion_criteria_review or sociodemographics does not exist?')
count=0
count2=0
for s in subjects:
    p=s[:2]+'-'+s[:9]+'-form_inclusionexclusion_criteria_review-day*to*csv'
    _p=s[:2]+'-'+s[:9]+'-form_sociodemographics-day*to*csv'

    p1=glob(p)
    _p1=glob(_p)

    if not (p1 and _p1):
        count+=1
    else:
        # both exists
        df=pd.read_csv(p1[0])
        _df=pd.read_csv(_p1[0])

        try:
            if pd.isna(df.loc[0,'included_excluded']) or pd.isna(df.loc[0,'chrcrit_part']) or pd.isna(_df.loc[0,'chrdemo_sexassigned']):
                pass

        except:
            # at least one of the three variables does not exist
            count2+=1

print(count)
print('for how many subjects at least one of the three variables does not exist?')
print(count2)