将 pandas dataframe 上传到谷歌电子表格

Question

I followed the steps here and here but couldn't upload a pandas dataframe to google sheets.

First I tried the following code:

import gspread
from google.oauth2.service_account import Credentials

scope = ['https://spreadsheets.google.com/feeds',
         'https://www.googleapis.com/auth/drive']

credentials = Credentials.from_service_account_file('my_json_file_name.json', scopes=scope)

gc = gspread.authorize(credentials)

spreadsheet_key = '1FNMkcPz3aLCaWIbrC51lgJyuDFhe2KEixTX1lsdUjOY'
wks_name = 'Sheet1'
d2g.upload(df_qrt, spreadsheet_key, wks_name, credentials=credentials, row_names=True)

The above code returns an error message like this: AttributeError: module 'df2gspread' has no attribute 'upload' which doesn't make sense since df2spread indeed has a function called upload.

Second, I tried to append my data to a dataframe that I artificially created on the google sheet by just entering the column names. This also didn't work and didn't provide any results.

import gspread_dataframe as gd

ws = gc.open("name_of_file").worksheet("Sheet1")
existing = gd.get_as_dataframe(ws)
updated = existing.append(df_qrt)
gd.set_with_dataframe(ws, updated)

Any help will be appreciated, thanks!

Answer 1

You are not importing the package properly.

Just do this

from df2gspread import df2gspread as d2g

When you convert a worksheet to Dataframe using

existing = gd.get_as_dataframe(ws)

All the blank columns and rows in the sheet are now part of the dataframe with values as NaN, so when you try to append it with another dataframe it won't be appended because columns are mismatched. Instead try this to covert worksheet to dataframe

existing = pd.DataFrame(ws.get_all_records())

When you export a dataframe in Google Sheets the index of the dataframe is stored in the first column(It happened in my case, can't be sure). If the first column is index then you can remove the column using

existing.drop([''],axis=1,inplace=True)

Then this will work properly.

updated = existing.append(df_qrt)
gd.set_with_dataframe(ws, updated)