How to read the data from txt to python pandas if the data looks compacted by date?-CodePudding

I received a text file, which looks like

Time Purchasing_item
20120701
00:00 Apple pie
00:01 Coca-cola
20120702
00:00 Pepsi
00:01 Fish-ball

However, I want when I load the data into pandas, I want the data to look like

Added one column myself, but don't want to duplicate the date so many times beforehand.

Date     Time  Purchasing_item

20120701 00:00 Apple pie
20120701 00:01 Coca-cola
20120702 00:00 Pepsi
20120702 00:01 Fish-ball

Or... Make 20120701 00:00 as a timestamp

Time   Purchasing_item

20120701 00:00 Apple pie
20120701 00:01 Coca-cola
20120702 00:00 Pepsi
20120702 00:01 Fish-ball

Could ask, how could I make one of the versions without duplicating the date many times manually?

Many thanks for your answer.

CodePudding user response：

Here is one way to do it:

df = pd.read_table('Data/file.txt', names=['DateTime'], skiprows=1)

df[['DateTime', 'Purchasing_item']] = df['DateTime'].str.split(n=1, expand=True)

for i in df.index:
    if df['Purchasing_item'].isna()[i] == True:
        df.iloc[i 1, 0] = df.iloc[i, 0]   ' '   df.iloc[i 1, 0]
        df.iloc[i 2, 0] = df.iloc[i, 0]   ' '   df.iloc[i 2, 0]

df.dropna(inplace=True)

Output:

   DateTime           Purchasing_item
1  20120701 00:00           Apple pie
2  20120701 00:01           Coca-cola
4  20120702 00:00               Pepsi
5  20120702 00:01           Fish-ball