def combine_with_rd_raw(df, nid, extract_type_id, location_set_version_id):
merge_cols = [
'nid', 'extract_type_id', 'location_id', 'year_id', 'age_group_id',
'sex_id', 'cause_id', 'site_id'
]
val_cols = ['deaths']
raw_df = get_phase_output("disaggregation",
nid=nid,
extract_type_id=extract_type_id)
raw_df = raw_df.groupby(merge_cols, as_index=False)[val_cols].sum()
raw_df = raw_df.rename(columns={'deaths': 'deaths_raw'})
corr_df = get_phase_output("misdiagnosiscorrection",
nid=nid,
extract_type_id=extract_type_id)
corr_df = corr_df.groupby(merge_cols, as_index=False)[val_cols].sum()
corr_df = corr_df.rename(columns={'deaths': 'deaths_corr'})
rd_df = get_phase_output("redistribution",
nid=nid,
extract_type_id=extract_type_id)
rd_df = rd_df.rename(columns={'deaths': 'deaths_rd'})
df = df.merge(rd_df, how='left', on=merge_cols)
df = df.merge(corr_df, how='left', on=merge_cols)
df = df.merge(raw_df, how='left', on=merge_cols)
for val_col in ['deaths_rd', 'deaths_corr', 'deaths_raw']:
df[val_col] = df[val_col].fillna(0)
return df
def main(nid, extract_type_id, code_system_id, launch_set_id):
"""Collect source metadata."""
print_log_message("Reading disaggregation data for source metadata phase.")
# in stata, this was pulled using corrections phase output (step 3)
# use data after all garbage codes have been set (step 3 produces ZZZ)
# but before any redistribution (including HIV redistribution)
df = get_phase_output("disaggregation", nid=nid,
extract_type_id=extract_type_id)
data_type_id = get_value_from_nid(nid, "data_type_id", extract_type_id)
source = get_value_from_nid(nid, "source", extract_type_id)
representative_id = get_value_from_nid(nid, "representative_id",
extract_type_id)
df = run_phase(df, nid, extract_type_id, data_type_id, source,
representative_id, code_system_id)
print_log_message(
"Writing {n} rows of output for launch set {ls}, nid {nid}, extract "
"{e}".format(n=len(df), ls=launch_set_id, e=extract_type_id, nid=nid)
)
write_phase_output(df, 'sourcemetadata', nid,
extract_type_id, launch_set_id)
def main(nid, extract_type_id, launch_set_id):
"""Read the data, run the phase, write the output."""
print_log_message("Reading redistribution data..")
df = get_phase_output('redistribution',
nid=nid,
extract_type_id=extract_type_id)
cause_set_version_id = int(CONF.get_id('cause_set_version'))
pop_run_id = int(CONF.get_id('pop_run'))
location_set_version_id = int(CONF.get_id('location_set_version'))
# run the phase
df = run_phase(df, nid, extract_type_id, pop_run_id, cause_set_version_id,
location_set_version_id)
# upload to database
print_log_message(
"Writing {n} rows of output for launch set {ls}, nid {n}, extract "
"{e}".format(n=len(df), ls=launch_set_id, e=extract_type_id))
write_phase_output(df, 'corrections', nid, extract_type_id, launch_set_id)