0

i want to read file name in folder which i already did using file=glob.glob... function. and add in 'file_last_mod_t' column last modify file time.

my part of code:

df=pd.DataFrame(columns=['filename','file_last_mod_t','else'])

df.set_index('filename')
for file in glob.glob('folder_path'): #inside this folder is file.txt
    file_name=os.path.basename('folder_path')
    df.loc[file_name]= os.path.getmtime(file)

which gives me:

df:
filename,file_last_mod_t,else
file.txt,123456,123456          #123456 its time result example

i want to add this last modify time only to file_last_mod_t column, not for all.

i want to receive :

df:
filename,file_last_mod_t,else
file.txt,123456,

thanks in advice

after code modification:

df=pd.read_csv('C:/df.csv')
filename_list= pd.Series(result_from_other_definition)# it looks same as in #filename column
df['filename']=filename_list # so now i have dataframe with 3 columns and firs column have files list
df.set_index('filename')
      for file in glob.glob('folder_path'):#inside this folder is file.txt
      df['file_last_mod_t']=df['filename'].apply(lambda x: (os.path.getmtime(x)) #the way how getmtime is present is now no matter, could be #float numbers
      df.to_csv('C:/df.csv')

#printing samples: first run:

df['filename']=filename_list
print (df)
,'filename','file_last_mod_t','else'
0,file1.txt,NaN,NaN
1,file2.txt,NaN,NaN

code above works fine after first run when df is empty , only with headers. after next run when i run the code and df.csv have some content i am changing manually value of timestamp in file, i am receiving an error : TypeError: stat: path should be string, bytes, os.PathLike or integer,not float this code should replace manually modified cell with good timestamp. i think it's connected with apply also i dont know why index appear in df

**solved **

sygneto
  • 1,761
  • 1
  • 13
  • 26

1 Answers1

1

Please see comment on code as following:

import os
import pandas as pd
import datetime as dt
import glob

# this is the function to get file time as string
def getmtime(x):
    x= dt.datetime.fromtimestamp(os.path.getmtime(x)).strftime("%Y-%m-%d %H:%M:%d")
    return x

df=pd.DataFrame(columns=['filename','file_last_mod_t','else'])

df.set_index('filename')

# I set filename list to df['filename']
df['filename'] = pd.Series([file for file in glob.glob('*')])

# I applied a time modified file to df['file_last_mod_t'] by getmtime function
df['file_last_mod_t'] = df['filename'].apply(lambda x: getmtime(x))

print (df)

The result is

          filename      file_last_mod_t else
0        dataframe  2019-05-04 18:43:04  NaN
1      fer2013.csv  2018-05-26 12:18:26  NaN
2         file.txt  2019-05-04 18:49:04  NaN
3        file2.txt  2019-05-04 18:51:04  NaN
4   Untitled.ipynb  2019-05-04 17:41:04  NaN
5  Untitled1.ipynb  2019-05-04 20:51:04  NaN

For the updated question, I started with df.csv what have data as following:

filename,file_last_mod_t,else
file1.txt,,

And, I think you want to add new files. So, I made the code as following:

import os
import pandas as pd

df=pd.read_csv('df.csv')

df_adding=pd.DataFrame(columns=['filename','file_last_mod_t','else'])
df_adding['filename'] = pd.Series(['file2.txt'])
df = df.append(df_adding)
df = df.drop_duplicates('filename')

df['file_last_mod_t']=df['filename'].apply(lambda x: (os.path.getmtime(x))) #the way how getmtime is present is now no matter, could be #float numbers
df.to_csv('df.csv', index=False)

I created df_adding dataframe for new files and I appended it to df which read df.csv. Finally, we can apply getmtime and save if to df.csv.

yaho cho
  • 1,779
  • 1
  • 7
  • 19
  • hello, df is empty, there is no filenames inside, only headers. I need to add filename to `df['filename]' and last modify time for this file , and sory i did mistake in case of reading DF please take a look on edits, thank You – sygneto May 04 '19 at 11:25
  • hello, could you tell me what function i need to use instead `apply` ? because i am receiving errors when some rows in `file_last_mod_t` already exist and this code cannot replace them with new values, i think;( error : `TypeError: stat: path should be string, bytes, os.PathLike or integer, not float` – sygneto May 07 '19 at 15:06
  • 1
    @sygneto could you give me your code and the result of `print(df)` before run `.apply`? I think you can add it on your question. – yaho cho May 08 '19 at 04:28
  • 1
    @sygneto Updated on my answer. But, I do not make sure the meaning of `have some content`. – yaho cho May 08 '19 at 07:45
  • its meand that after first run df.csv have information inside, thanks – sygneto May 08 '19 at 07:49
  • i think error is becaouse of thiss line : `df['filename']=filename_list` it cannot add filenames which are already there, maybe i should try , `.replace`? – sygneto May 08 '19 at 08:31
  • 1
    @sygneto Do you want to add new files from `filename_list= pd.Series(result_from_other_definition)`? What is `result_from_others_definition`? Is it a list of new files? – yaho cho May 08 '19 at 08:42
  • Yes, it is list of new files – sygneto May 08 '19 at 08:59
  • thanks its much better, THANK YOU ! i used .drop_duplicates() because i had some when filename was already there, cheers! – sygneto May 08 '19 at 09:36
  • @sygneto Yeap. I updated `drop_duplicates()` on answer too. Have a nice day~! – yaho cho May 08 '19 at 09:42