Consider allowing option to drop duplicates in set_with_dataframe about gspread-dataframe

[robin900]

I do not understand what behavior you desire. Are you saying that you would like to skip over two columns of the dataframe when writing to the worksheet? But also somehow skipping over two columns in the worksheet so that those two columns in the sheet are unaffected by the dataframe cells being written?

[robin900]

Just guessing here, but if you only wish to write some column or columns of a dataframe to a worksheet, derive a second dataframe that contains just the column or columns to write; then use parameters for set_as_dataframe that allow you to offset the column, etc.:

def set_with_dataframe(
    worksheet,
    dataframe,
    row=1,
    col=1,
    include_index=False,
    include_column_header=True,
    resize=False,
    allow_formulas=True,
    string_escaping="default"):

[MrBeardedGuy]

My bad. Too little description :)
I'm thinking of a scenario where a dataframe gets appended to a worksheet with some data in it. The action might result in duplicated cols or rows that need to be dropped (impossible to predict which ones though).

Appending can be easily done by passing the option row=last_row_of_sheet_data+1 to the set_with_dataframe

I was wondering if set_with_dataframe could have an option to remove the duplicated values in case the worksheet already has some data. Similar to what Panda's drop duplicates does.

[robin900]

The most successful approach in using the gspread-dataframe package is:

1. Read the contents of a worksheet into a DataFrame.
2. Modify the DataFrame as desired. Add data, drop duplicates, remove columns, etc.
3. Write the resulting DataFrame to a worksheet, replacing the entire contents of the worksheet.

I expect that this approach will serve your DataFrame writing needs. If you encounter in the future a specific situation where the above approach isn't sufficient, feel free to open another issue.