[Python] How to set a boolean flag True after entry and back to False after exit using Pandas...

I'm working with a Pandas DataFrame where I need to track a boolean flag (in_position)
that becomes True after an entry_signal and resets to False only after an exit_signal.

This must be done vectorized only — no loops — to support millions of rows efficiently.

✅ Expected behavior:

• in_position should turn True only when entering a new state (after a clean entry)
• It should stay True, even if additional entry_signals appear inside that position
• It should return to False only after a valid exit_signal

In other words, entry_signal values during an open position must be ignored.

❌ The problem:


The issue seems to come from the way cumulative comparison works:

df['in_position'] = df['final_entry'].cumsum() > df['final_exit'].cumsum()


Even when both cumulative values reach 1, this comparison still evaluates to True on the row after the final_exit, because both values are equal, and > is no longer satisfied only on the same row, but not before it.

This results in in_position remaining True until the next row — or forever if no new exit appears

I understand that this is a side effect of how .cumsum() and > behave in vectorized logic.
What I'm struggling with is:
➡️ How to replace this logic with a vectorized alternative that truly toggles the position on entry and resets it exactly at the exit?

Any help would be appreciated!

Minimal reproducible example:

# Step 1: Create minimal dataset with relevant values
data = {
'Close': [110, 111, 112, 113, 114, 115, 116],
'RSI_10': [None, 20, None, 41, 20, None, 61],
'SMA_200': [100]*7,
}
index = pd.to_datetime([
'2025-04-21', '2025-04-22', '2025-04-23',
'2025-04-24', '2025-04-25', '2025-04-27', '2025-04-28'
])
df = pd.DataFrame(data, index=index)

# Step 2: Signals
df['entry_signal'] = (df['RSI_10'] < 30) & (df['Close'] > df['SMA_200'])
df['partial_exit_signal'] = df['RSI_10'] > 40
df['exit_signal'] = df['RSI_10'] > 60

# Step 3: Cumulative logic - the main problem
entry_cumsum = df['entry_signal'].cumsum()
exit_cumsum = df['exit_signal'].cumsum()

in_position = entry_cumsum > exit_cumsum
in_position_prev = in_position.shift(1).fillna(False).astype(bool)

# Step 4: Final signals
df['final_entry'] = df['entry_signal'] & ~in_position_prev
df['partial_exit'] = df['partial_exit_signal'] & in_position_prev
df['final_exit'] = df['exit_signal'] & in_position_prev
df['in_position'] = in_position
df['in_position_prev'] = in_position_prev

print(df[['RSI_10', 'Close', 'entry_signal', 'final_entry', 'final_exit', 'in_position']])

Continue reading...