Version 1.19.0

I am getting this error when using edit dataframe, someone please help me

Thanks. I need this feature very much. At present, I have to use st.columns to make particularly ugly table with widgets. Can you give an approximate release time? Like weeks or months? Or follow the next release?

We already fixed this issue a few days ago and it will be released with 1.20 in a few days. In the meantime, you might be able to fix this by updating numpy.

1 Like

Thank you very much. I updated numpy to the latest but it still doesn’t solve the problem. I will wait for the new update and hope it will solve my problem

Great and useful feature! But can’t it support pd.styler? For example, st.dataframe(
can create a table with highlighted blocks (green background color), while
cannot show the highlighted color.

Unfortunately, supporting pandas styler would seriously slow down the data editor. But we exploring alternative ways to apply styling and modify display values.

Awesome! Thanks for the effort!

1 Like

Not having styler is disappointing news in the editor is disappointing news for my scenario. I have a dataframe with styling on several columns, which I do not expect to be editable, while there are a few other columns that I would like to make editable. These wouldn’t necessarily be styled, except possibly for number formatting. Would that still cause trouble with performance? Thanks.

1 Like

I think this should be possible to activate styler support for non-editable columns. We will evaluate this once we have column config released (that will make it possible to make specific columns non-editable)

1 Like

Thank you for the information. I also work with data frames, and your solution works. However, set_index causes a problem in some cases. Thus, I use the following:

df.index = new_index_column

It works for all data frames. Also, use_container_width=True solves my 2nd issue. Thanks again!

Hello, I love this experimental_data_editor, but I observe that the line deletion or addition often needs to be done twice to be visible. Is it something that other also observed, that would suggest it’s a bug from the tool ? Or is there 100% chance the issue comes from my code ?