St.dataframe displays int as float

[rimano]

Summary

I am trying to display a pandas dataframe but the columns containing integers are displayed as floats. (picture below)

The columns NPDCode and Year are supposed to be displayed as integers.

Expected behavior:

Here’s what I get when using the same function in jupyter notebook (in vscode)

Debug info

• Streamlit version: latest
• Python version: 3.11.1
• OS version: Windows 10

[Franky1]

@rimano

integers are displayed as floats

It looks more like it is the common way of representing thousands in English, as you can also see further to the right.

[rimano]

Hi @Franky1 ! Thank you for your reply.

I get your point but is there a way of preventing that behaviour?

[ac3]

df[‘Year’]=df[‘Year’].astype(str)

[rimano]

Hi @ac3! Thank you for your reply.

I can’t really do that as I use the same dataframe later to plot the annual production (picture below). The x axis cannot be made of strings.

Note: The original dataset came unsorted. To sort it, I need to keep that column’s dtype as integer.

[Franky1]

@rimano

I can’t really do that as I use the same dataframe later

You could split up the dataframe into two dataframes, one for the table display and one for further numerical analysis. Not ideal, but a workaround.

[harshhh]

@rimano You can use st.table(df)

Using st.table(df)

Using st.dataframe(df)

The difference is st.table() will show df as static table and st.dataframe() allow you to sort data.

[Bwhiz]

If displaying the dataframe is important (as I assume it is), then make a copy of the dataframe (using df.copy()) set it to a new variable, cast the dtype of this new variable to integer and display your plots from the new table.
This should help

[atriumsplotch]

Had the same issue seen numbers like this: 120,131,009,003,829.98 instead of like this: 0120131009003830 same as in .csv file.
To solve this problem I used pandas style to specify “commas” to separate thousands.

df.style.format(thousands=',')

I just removed comma between ' symbols to make it solid number :

st.dataframe(df.style.format(thousands=''))

[rimano]

Hi @atriumsplotch ! Thank you for your reply.

Your solution worked like a charm. However, I ended up doing something slightly different. I used the precision parameter as well.

Indeed, using the method you suggested led to having a bunch of unwanted zeros at the end (apparently, the default is 6 decimals).

styler.format.precision: default 6.

So I did this:

df.style.format(precision=0, thousands='')

And If needed, you can always get the full value by double clicking on the cell.

Result

Honestly, I didn’t even know you could style a dataframe so thanks again .