Filled_qty doesn't balance at end of day

I’m doing paper trading where at the start and end of every day the portfolio has 0 shares. Yet, when I execute the following code (today is the 24’th), the sum total of shares bought - sold doesn’t end up 0 as exemplified by the output:

False
26893
from load_alpaca_apis import api
import pandas as pd
after=None
while True:
    filled_orders = api.list_orders(status="closed", after=after)
    if len(filled_orders) == 0:
        break
    filled_orders = [l for l in filled_orders if l.status=='filled']
    df = pd.DataFrame(data=[r._raw for r in filled_orders], columns=filled_orders[0]._raw.keys())
    try: 
        filled_orders_df = pd.concat([filled_orders_df, df])
    except:
        filled_orders_df = df
    after = filled_orders_df.submitted_at.iloc[-1]

filled_orders_df['signed_side'] = [1 if x == 'buy' else -1 for x in filled_orders_df.side]
filled_orders_df.filled_qty = [eval(x) for x in filled_orders_df.filled_qty]
filled_orders_df['signed_qty'] = filled_orders_df.filled_qty * filled_orders_df.signed_side
filled_orders_df.to_csv("orders.csv")
df = pd.read_csv("orders.csv", parse_dates=['filled_at'], infer_datetime_format=True).set_index("filled_at")
print(any([x.day!=24 for x in df.index]))
print(df.signed_qty.sum())