Datatable datetime format

Dash Python
1

I have data from a pandas dataframe that I want to put in a datatable. The first field is a date. It’s formatted YYYY-MM-DD in pandas. In the dataframe it’s format is YYYY-MM-DDT00:00:00. How do I get time out of the format in the dataframe? In all of the examples in the DataTable user guide dates are in YYYY-MM-DD format.

This worked properly in an earlier version of Dash and DataTable (I believe the version of Dash was 0.4).

2

I’m also having this issue. Anyone found a fix?

3

I hit all my date columns with this before populating the DataTable component:

for x in ["Date Col1", "Date Col2", "Date Col3"]:
        if x in df.columns:
            df[x] = pd.to_datetime(df[x], format="%Y-%m-%d").dt.floor("d")
1 Like
4

That doesn’t work for me unfortunately, Dash still adds the extra ‘00:00:00’.

5

Assuming that your column is named as “Date_Col”, you can format it before creating the data table:

df.Date_Col = pd.DatetimeIndex(df.Date_Col).strftime("%Y-%m-%d")

That should do it.

Also refer to this documentation for further details https://dash.plot.ly/datatable/typing.

1 Like
6

Thanks that worked! I’ll check out that documentation

7

Sorting doesn’t work since strftime converts to string

8

I haven´t yet found a ‘format’ solution in the table that allows to sort by date.

Workaround:
I used the index numbers of the df (sorted by datetime) as an “id” column in the table to sort by date.

9

Just bumping this topic. I’m not understanding how I can format a datetime column while still allowing it to be sorted correctly.

Something like:

columns=[{'id':'Date','name':'My Formatted Date','type': 'datetime', 'format:'???'}]

But I’m not understanding what that ??? should be.

10

According to https://dash.plotly.com/datatable/reference, columns.format is “derived from the d3-format library specification”, which doesn’t have any date-formatting functionality. So that definitely won’t work. If you change the date column of the DataFrame into a string with the format YYYY-MM-DD, before you send the DataFrame to the Datatable, then it will display correctly and sort correctly, because alphabetical order on the string is the same as chronological order.

1 Like
11

For me the way it works best is to convert the data in pandas to dt.date format:
df[date_col] = df[date_col].dt.date

When you move this to datatable it keeps just YYYY-MM-DD with no time and it still is a date.