2345ff2a2a1a3da295890b19e453a250dfbd3baf,pythalesians/economics/events/eventstudy.py,EventStudy,get_surprise_against_intraday_moves_over_custom_event,#EventStudy#Any#Any#Any#Any#Any#Any#Any#Any#Any#Any#Any#,144

Before Change

        ticker = event_fx + "-" + event_name + ".release-date-time-full"

        data_frame_agg = None
        data_frame_cross_orig = data_frame_cross_orig.resample("T")         // resample to minute freq - in case there are missing values

        ef_time_start = ef_time_frame[ticker] - timedelta(minutes=1)        // start time
        indices_start = data_frame_cross_orig.index.isin(ef_time_start)

        for offset in offset_list:
            data_frame_cross = data_frame_cross_orig

            ef_time = ef_time_frame[ticker] + timedelta(minutes=offset - 1)     // end time

            // calculate returns over the x min event
            indices = data_frame_cross.index.isin(ef_time)
            col_dates = data_frame_cross.index[indices]

            col_rets = (data_frame_cross.iloc[indices].values) \
                       / (data_frame_cross.iloc[indices_start].values) - 1

            mkt_moves = pandas.DataFrame(index=col_dates)
            mkt_moves[cross + " " + str(offset) + "m move"] = col_rets
            mkt_moves.index.name = ticker
            mkt_moves.index = col_dates - timedelta(minutes=offset - 1)

            data_frame = ef_time_frame.join(mkt_moves, on=ticker, how="inner")
            temp_index = data_frame[ticker]

            spot_moves_list = []

            if data_frame_agg is None:
                data_frame_agg = data_frame
            else:
                label = cross + " " + str(offset) + "m move"
                spot_moves_list.append(label)
                data_frame = data_frame[label]
                data_frame.index = temp_index
                data_frame_agg = data_frame_agg.join(data_frame, on=ticker, how="inner")

        if add_surprise == True:
            data_frame_agg[event_fx + "-" + event_name + ".surprise"] = data_frame_agg[event_fx + "-" + event_name + ".actual-release"] \
                               - data_frame_agg[event_fx + "-" + event_name + "." + surprise_field]

        return data_frame_agg

After Change

        ticker = event_fx + "-" + event_name + ".release-date-time-full"

        data_frame_agg = None
        data_frame_cross_orig = data_frame_cross_orig.resample("T").mean()  // resample to minute freq - in case there are missing values

        ef_time_start = ef_time_frame[ticker] - timedelta(minutes=1)        // start time
        indices_start = data_frame_cross_orig.index.isin(ef_time_start)

        for offset in offset_list:
            data_frame_cross = data_frame_cross_orig

            ef_time = ef_time_frame[ticker] + timedelta(minutes=offset - 1)     // end time

            // calculate returns over the x min event
            indices = data_frame_cross.index.isin(ef_time)
            col_dates = data_frame_cross.index[indices]

            col_rets = (data_frame_cross.iloc[indices].values) \
                       / (data_frame_cross.iloc[indices_start].values) - 1

            mkt_moves = pandas.DataFrame(index=col_dates)
            mkt_moves[cross + " " + str(offset) + "m move"] = col_rets
            mkt_moves.index.name = ticker
            mkt_moves.index = col_dates - timedelta(minutes=offset - 1)

            data_frame = ef_time_frame.join(mkt_moves, on=ticker, how="inner")
            temp_index = data_frame[ticker]

            spot_moves_list = []

            if data_frame_agg is None:
                data_frame_agg = data_frame
            else:
                label = cross + " " + str(offset) + "m move"
                spot_moves_list.append(label)
                data_frame = data_frame[label]
                data_frame.index = temp_index
                data_frame_agg = data_frame_agg.join(data_frame, on=ticker, how="inner")

        if add_surprise == True:
            data_frame_agg[event_fx + "-" + event_name + ".surprise"] = data_frame_agg[event_fx + "-" + event_name + ".actual-release"] \
                               - data_frame_agg[event_fx + "-" + event_name + "." + surprise_field]

        return data_frame_agg

Instances