def data_for(data_date, series, earliest, pct: bool):
bars = Bars.get(series, earliest=earliest)
data, data_date = bars.data_for(data_date)
if pct:
data = data.divide(data.sum(axis='columns'), axis='rows')
data = data.rolling(7).mean().dropna(how='all')
return data, data_date, bars
def main():
parser = ArgumentParser()
add_date_arg(parser, help='first release to use', default=earliest_testing)
add_parallel_args(parser,
default_duration=0.1,
default_output='gif',
from_date=False)
parser.add_argument('config', choices=BARS.keys())
add_date_arg(parser,
'--earliest',
help='min for x-axis',
default=second_wave)
parser.add_argument('--diff-log-scale', action='store_true')
parser.add_argument('--diff-no-lims', action='store_true')
parser.add_argument('--ylim', type=float)
parser.add_argument('--y-max-factor', type=float, default=1.02)
args = parser.parse_args()
config = Bars.get(args.config, earliest=args.earliest)
from_date = max(args.from_date, args.earliest)
dates = available_dates(config.metric,
config.data_file_stem,
earliest=from_date)
to_date = parallel_to_date(args, dates[0])
if to_date != dates[0]:
dates = [d for d in dates if d <= to_date]
latest_date = dates[0]
max_metric = config.data_for(latest_date)[0].sum(
axis=1).max() * args.y_max_factor
params = {}
if args.diff_no_lims:
params['diff_ylims'] = None
lines = config.lines_for(latest_date)
if lines:
params['line_ylim'] = max(l.data.max()
for l in lines) * args.y_max_factor
parallel_render(
f'animated_{args.config}',
partial(plot_bars,
config=config,
ylim=args.ylim or max_metric,
diff_log_scale=args.diff_log_scale,
show_title=True,
to_date=date.today(),
**params), dates,
**parallel_params(args, item_is_timestamp=False))