Allow us to set how uploaded data should be interpreted.

[Ctilhs]

From what I understand, by default when we use the upload() from df2gspread, dataframe values are converted to string then upload is performed.

I struggled with it when I had to upload values which were prices. Same situation if I would like them to stay as digits after update, without my interaction in google docs.

Simple fix.
In upload() add argument value_input_opt='RAW'

df2gspread/df2gspread/df2gspread.py

Lines 24 to 26 in 01bc46d

	def upload(df, gfile="/New Spreadsheet", wks_name=None,
	col_names=True, row_names=True, clean=True, credentials=None,
	start_cell = 'A1', df_size = False, new_sheet_dimensions = (1000,100)):

And in three function calls value_input_option=value_input_opt

df2gspread/df2gspread/df2gspread.py

Line 126 in 01bc46d

wks.update_cells(cell_list)

df2gspread/df2gspread/df2gspread.py

Line 134 in 01bc46d

wks.update_cells(cell_list)

df2gspread/df2gspread/df2gspread.py

Line 147 in 01bc46d

wks.update_cells(cell_list)

From: developers.google.com
What we get? Now in upload we could set parameter value_input_option='USER_ENTERED', so when we upload digits they will be automatically parsed as numbers.
If we will not specify value_input_option in upload() it will stay default as 'RAW' so data will be uploaded as it is now.

Not sure is it related?: #41

It's my first time ever adding Issue at github. I will gladly accept advice if I should improve something.

[saitx]

Yeah for now I have been having a duplicate sheet with =VALUE(Raw_Sheet!H2) and so on. This would be a great addition. Something like a dtype parameters that pandas has would be great.

[maybelinot]

@Ctilhs thank you for creating an issue, that would be great if you could submit a PR with proposed changes. Otherwise I'll work on this as soon as I'll have time.

[winf-hsos]

Any updates on this?

[varunjain3]

Are the authors alive? I wanted this as well, shall I create a pr?

[maybelinot]

Hi @varunjain3 I don't actively support the lib anymore, if you will open a pr with tests - I'll review and bump the lib version

[varunjain3]

Thanks for the prompt reply, I’ll do the needful, also are there any alternatives you found for the same project?

[maybelinot]

Not really, df2gspread mainly leverages https://github.com/burnash/gspread, which is update regularly from what I can see, so you may built small abstraction layer to fulfil your needs just on top of that.

[mroy-seedbox]

I'd suggest abandoning (that's what we did), when an entire library can be replaced with a couple lines of monkey patching in the actively maintained gspread lib:

    def upload_pandas_df(self, df):
      values = [df.columns.values.tolist()]
      values.extend(df.values.tolist())
      sheet.values_update(
        self.title,
        params = { 'valueInputOption': 'USER_ENTERED' },
        body = { 'values': values }
      )

    gspread.Worksheet.upload_pandas_df = upload_pandas_df

Or use gspread-dataframe instead.