Python build_agg_meta Examples

Example #1

File: revert_rate.py Project: wikimedia/analytics-user-metrics

# ==========================
# DEFINE METRIC AGGREGATORS
# ==========================

metric_header = RevertRate.header()

field_prefixes = \
    {
        'reverted_': 1,
        'reverts_': 2,
        'revisions_': 3,
    }

# Build "stats" aggregator
op_list = [sum, mean]
revert_stats_agg = build_numpy_op_agg(build_agg_meta(op_list, field_prefixes),
                                      metric_header,
                                      'revert_stats_agg')

agg_kwargs = getattr(revert_stats_agg, METRIC_AGG_METHOD_KWARGS)
setattr(revert_stats_agg, METRIC_AGG_METHOD_KWARGS, agg_kwargs)

# Build proportion aggregator
revert_prop_agg = boolean_rate
revert_prop_agg = decorator_builder(RevertRate.header())(
    revert_prop_agg)

setattr(revert_prop_agg, METRIC_AGG_METHOD_FLAG, True)
setattr(revert_prop_agg, METRIC_AGG_METHOD_NAME, 'revert_prop_agg')
setattr(revert_prop_agg, METRIC_AGG_METHOD_HEAD, ['total_users',
                                                     'total_reverted',

Example #2

File: bytes_added.py Project: dartar/user_metrics

# DEFINE METRIC AGGREGATORS
# ==========================

metric_header = BytesAdded.header()

field_prefixes = {
    'net_': 1,
    'abs_': 2,
    'pos_': 3,
    'neg_': 4,
    'count_': 5,
}


# Build "mean" decorator
ba_mean_agg = build_numpy_op_agg(build_agg_meta([mean], field_prefixes),
                                   metric_header, 'ba_mean_agg')
# Build "standard deviation" decorator
ba_std_agg = build_numpy_op_agg(build_agg_meta([std], field_prefixes),
                                   metric_header, 'ba_std_agg')
# Build "sum" decorator
ba_sum_agg = build_numpy_op_agg(build_agg_meta([sum], field_prefixes),
                                  metric_header, 'ba_sum_agg')
# Build "median" decorator
ba_median_agg = build_numpy_op_agg(build_agg_meta([median], field_prefixes),
                                   metric_header, 'ba_median_agg')
# Build "min" decorator
ba_min_agg = build_numpy_op_agg(build_agg_meta([min], field_prefixes),
                                metric_header, 'ba_min_agg')
# Build "max" decorator
ba_max_agg = build_numpy_op_agg(build_agg_meta([max], field_prefixes),

Example #3

File: pages_created.py Project: kl07/wikipedia_user_metrics

    if metric_params.log_:
        logging.info(__name__ + '::Processed PID = %s.  '
                                'Dropped users = %s.' % (
                                    os.getpid(), str(dropped_users)))

    return results


# ==========================
# DEFINE METRIC AGGREGATORS
# ==========================

from user_metrics.etl.aggregator import build_numpy_op_agg, build_agg_meta
from user_metrics.metrics.user_metric import METRIC_AGG_METHOD_KWARGS

metric_header = PagesCreated.header()

field_prefixes =\
    {
        'count_': 1,
    }

# Build "dist" decorator
op_list = [sum, mean, std, median, min, max]
pages_created_stats_agg = build_numpy_op_agg(
    build_agg_meta(op_list, field_prefixes), metric_header,
    'pages_created_stats_agg')

agg_kwargs = getattr(pages_created_stats_agg, METRIC_AGG_METHOD_KWARGS)
setattr(pages_created_stats_agg, METRIC_AGG_METHOD_KWARGS, agg_kwargs)

Example #4

File: time_to_threshold.py Project: sudeepdas/E3_analysis

# ==========================
# DEFINE METRIC AGGREGATORS
# ==========================

# Build "average" aggregator
ttt_avg_agg = weighted_rate
ttt_avg_agg = decorator_builder(TimeToThreshold.header())(ttt_avg_agg)

setattr(ttt_avg_agg, um.METRIC_AGG_METHOD_FLAG, True)
setattr(ttt_avg_agg, um.METRIC_AGG_METHOD_NAME, "ttt_avg_agg")
setattr(ttt_avg_agg, um.METRIC_AGG_METHOD_HEAD, ["total_users", "total_weight", "average"])
setattr(ttt_avg_agg, um.METRIC_AGG_METHOD_KWARGS, {"val_idx": 1})


metric_header = TimeToThreshold.header()

field_prefixes = {"time_diff_": 1}

# Build "dist" decorator
op_list = [median, min, max]
ttt_stats_agg = build_numpy_op_agg(build_agg_meta(op_list, field_prefixes), metric_header, "ttt_stats_agg")


if __name__ == "__main__":
    for i in TimeToThreshold(threshold_type_class="edit_count_threshold", first_edit=0, threshold_edit=1).process(
        [13234584]
    ):
        print i