Using dataframe_explorer to filter dataframe which has datetime object values

[saitharun_gunasekar]

Summary

Iam using dataframe_explorer from streamlit_extras Module for filtering a dataframe which contains a datetime object that is formatted (“%d/%m/%Y %H:%M:%S”) but when i run the code the time columns is formatted to (“%Y/%m/%d %H:%M:%S”)
and the filter options are also working properly , i even tried to converting into a pandas datetime object, df1[“Time”]=pd.to_datetime(filtered_df1[“Time”], format=“%d/%m/%Y %H:%M:%S”) and when i give input in the filter column the last row is not visible unless in reload the code

Steps to reproduce

@st.cache_data()
def update_data(_df1,_time_formatted,_count,_component_names,column_names):
_df1.loc[len(_df1)] = [_time_formatted, _count, _component_names]
def filter(column_names):
with st.sidebar:
st.subheader(“Filter Service Down Time History”)
filtered_df1 = dataframe_explorer(df1)
return filtered_df1
if name =“main”:
time1 = datetime.datetime.now()
time_formatted = time1.strftime(“%d/%m/%Y %H:%M:%S”)
update_data(df1,time_formatted, len(inactive_components), component_names, column_names)
df1[“Time”] = pd.to_datetime(df1[“Time”], format=“%d/%m/%Y %H:%M:%S”)
filtered_df, filtered_df1 = filter(column_names)

Expected behavior:
i want to be able to filter the data with the Time column and there shouldnt be any rows missing

Actual behavior:
when i run the code the time columns is formatted to (“%Y/%m/%d %H:%M:%S”)
and the filter options are also working properly and when i give input in the filter column the last row is not visible unless in reload the code

Debug info

• Streamlit version: Streamlit, version 1.22.0
• Python version: Python 3.11.2
• Using PyEnv - Pycharm
• OS version:
• Browser version:

[dataprofessor]

Hi @saitharun_gunasekar

Thanks for your question. It may be helpful to the community if the code snippet is a working example that reproduces the problem that you are encountering. Sharing the app’s code in its entirety (with library imports and the corresponding data used by the app) would definitely help.

Based on provided information, here’s what I would do to approach this problem:

• Check to see if deleting @st.cache_data() would resolve the missing rows. Thus, experiment with the placement of the cache statement so that it caches static data while not caching dynamically updated data.
• In regards to the date-time formatting, perhaps you can post-process the filtered dataframe to the desired date-time format

Hope this helps!

Best regards,
Chanin