Function timed_processing

test.py:146–162

(time_mins=360)

def timed_processing(time_mins=360):

    df = pd.read_csv('all_data.csv')  

    df['tickers'] = df.apply(lambda row: process_tickers(row.body), axis=1)
    df["ents"], df["sentiment"] = zip(*df.apply(nlp_processing, axis=1))
    df['created'] = df['created'].apply(lambda created: pd.Timestamp(created, unit='s')) 

    data_df = df.filter(['tickers', 'score', 'sentiment', 'created'])
    data_df['score'] = data_df.score.astype(np.int32)

    ex_tickers = data_df.explode('tickers').dropna()
    ex_tickers['sentiment'] = ex_tickers['sentiment'].apply(lambda row: row['compound'])
    ex_tickers['count'] = 1
    tks_group_6hours = ex_tickers.groupby(['tickers', pd.Grouper(key='created', freq=f'{time_mins}Min')])
    timed_df = pd.concat([tks_group_6hours['count'].count(), tks_group_6hours['sentiment'].mean(), tks_group_6hours['score'].sum()], axis=1)
    return timed_df.astype(np.float32)