如何通过 input() 在 pandas Dataframe 中按内容查找索引？

Question

I am trying to find an index based on the content of a table, which in this case is a string in the form of a date 'd/m/YYYY' or 'dd/mm/YYYY' for 2-digit days and months respectively. I am looking for the 1st of a specific month which i want to find interactively using input().

The code works fine as long as i hardcode the month:

stream = open(sys.argv[1], 'r')
for path in stream:
    df = pd.read_csv('{}'.format(path))
    df['Idx'] = df['Date'].str.contains('^1/8/20\d{2} 0:00:00')
    index_month = df[df['Idx']==True].index.values.astype(int)[0]
    print('Index was found at {}'.format(month, index_month))

Output:

Index was fount at 19516
Index was found at 19527
...
...

but gives

ValueError: cannot switch from automatic field numbering to manual field specification

when i use input() to find the respective month (see code below)

print("Enter Month as number:")
month=input()

stream = open(sys.argv[1], 'r')
for path in stream:
    df = pd.read_csv('{}'.format(path))
    df['Idx'] = df['Date'].str.contains('^1/{}/20\d{2} 0:00:00'.format(month))
    index_month = df[df['Idx']==True].index.values.astype(int)[0]
    print('Index was found at {}'.format(month, index_month))

The error is traced back to

df['Idx'] = df['Date'].str.contains('^1/{}/20\d{2} 0:00:00'.format(month))

Is it possible to solve it this way, or is there a better approach?

Answer 1

Your problem is simply here:

'^1/{}/20\d{2} 0:00:00'

{2} is interpreted by str's format method, so you need to escape it using double braces:

>>> month = 3
>>> re.search('^1/{}/20\d{{2}} 0:00:00'.format(month), '1/3/2020 0:00:00').group()
'1/3/2020 0:00:00'