[Python] Streamlit data_editor won't update data

I am trying to have a data_editor in streamlit that can handle automatic changes from users. The idea, is that user will update a column value and automatically the change will be displayed triggering a calculation within the table. In my example, it should days to expire for a financial option. Pretty much like excel. But I am stuck with the fact that, it will always revert to the previous and it needs "two clicks".

Here is the code I have (with a lot of chatpt too) At the moment I have creating the edited data frame, because it is failing to update the same data_editor. Ideally, the changes you see on the second data frame, I would like to appear in the input one.

Can anyone helpout?

In the code below, is where the problem is, user's input is not registered

import streamlit as st
import pandas as pd

contracts = ['Jan25', 'Feb25', 'Mar25','Apr25','May25','Jun25',
'Jul25', 'Auh25', 'Sep25', 'Oct25', 'Nov25', 'Dec25']
options_type = ['Call', 'Put']

# Only initialize once
if "editor_df" not in st.session_state:
st.session_state["editor_df"] = pd.DataFrame({
"Is Long": [True],
"Call/Put": [options_type[0]],
"Contract": [contracts[0]],
"Strike": [35.0]
})

# Show editor and get edited DataFrame
edited_df = st.data_editor(
st.session_state["editor_df"],
key="editor",
column_config={
"Is Long": st.column_config.CheckboxColumn(default=True),
"Call/Put": st.column_config.SelectboxColumn(options=options_type),
"Contract": st.column_config.SelectboxColumn(options=contracts),
"Strike": st.column_config.NumberColumn(default=35.0),
},
num_rows="dynamic"
)

# Compare before assigning
if not edited_df.equals(st.session_state["editor_df"]):
st.session_state["editor_df"] = edited_df.copy()


The below code is the workaround I have. Populate two tables. The first is the input and the second calculates the extra field.

import streamlit as st
import pandas as pd
import datetime

# Define contracts and expiries
contracts = ['Jan25', 'Feb25', 'Mar25', 'Apr25', 'May25', 'Jun25',
'Jul25', 'Aug25', 'Sep25', 'Oct25', 'Nov25', 'Dec25']
options_type = ['Call', 'Put']

expiries = {
'Jan25': datetime.datetime(2024, 12, 27),
'Feb25': datetime.datetime(2025, 1, 29),
'Mar25': datetime.datetime(2025, 2, 26),
'Apr25': datetime.datetime(2025, 3, 27),
'May25': datetime.datetime(2025, 4, 28),
'Jun25': datetime.datetime(2025, 5, 28),
'Jul25': datetime.datetime(2025, 6, 26),
'Aug25': datetime.datetime(2025, 7, 29),
'Sep25': datetime.datetime(2025, 8, 27),
'Oct25': datetime.datetime(2025, 9, 26),
'Nov25': datetime.datetime(2025, 10, 29),
'Dec25': datetime.datetime(2025, 11, 27)
}

# Initialize session state
if "editor_df" not in st.session_state:
st.session_state.editor_df = pd.DataFrame({
"Is Long": [True],
"Call/Put": ["Call"],
"Contract": ["Jan25"],
"Strike": [35.0]
})

# Show editor first
edited_df = st.data_editor(
st.session_state.editor_df,
key="editor",
column_config={
"Is Long": st.column_config.CheckboxColumn(),
"Call/Put": st.column_config.SelectboxColumn(options=options_type),
"Contract": st.column_config.SelectboxColumn(options=contracts),
"Strike": st.column_config.NumberColumn(),
},
num_rows="dynamic"
)

# Compute days on what user just changed
def compute_days(df: pd.DataFrame) -> pd.DataFrame:
now = datetime.datetime.now()

def get_days(contract):
expiry = expiries.get(str(contract))
if not isinstance(expiry, datetime.datetime):
return 0.0
delta = expiry - now
return max(delta.total_seconds() / 86400, 0.0)

df = df.copy()
df["Days to expiry"] = df["Contract"].apply(get_days)
return df

# Compute on latest edited data
final_df = compute_days(edited_df)

# Show full live-updated table
st.dataframe(final_df)

Continue reading...