Skip to content

Commit

Permalink
test ops oeb
Browse files Browse the repository at this point in the history
  • Loading branch information
@juliaGrandhay
juliaGrandhay committed Oct 28, 2024
1 parent 2116a5a commit 6de65ce
Showing 1 changed file with 104 additions and 93 deletions.
197 changes: 104 additions & 93 deletions patstat/p08b_outils_inpi_adress_api.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -1935,13 +1935,19 @@ def req_ops_oeb(df: pd.DataFrame) -> pd.DataFrame:

pt_ops2 = pd.concat(lpart, ignore_index=True)
pt_ops2["sequence"] = pt_ops2["sequence"].astype(int)
pt_ops2 = pt_ops2[["sequence", "name", "@change-date", "@change-gazette-num", "address", "pn", "type_party",
"@change-date2"]].drop_duplicates().reset_index(drop=True)
pt_ops2 = pt_ops2[["sequence", "address", "pn", "type_party"]].drop_duplicates().reset_index(drop=True)
pt_ops2["address_complete_fr"] = pt_ops2["address"].copy()

pt_ops3 = pd.merge(
df.drop(columns=["address_complete_fr", "com_code", "ville", "dep_id", "dep_nom", "reg_nom"]),
pt_ops2,
on=["pn", "type_party", "sequence"], how="inner")

pt_ops4 = adresses_ops_oeb(pt_ops3)

logger.info("end query OPS OEB")

return pt_ops2
return pt_ops4


@retry(tries=20, delay=5, backoff=5)
Expand Down Expand Up @@ -2383,93 +2389,98 @@ def create_df_address():
pubon = list(pat_oeb["pn"].unique())
print(f"Nombre de numéros de publication à chercher : {len(pubon)}.", flush=True)

df_pubon = pd.DataFrame(data={"pn": pubon})
sub_pubon = subset_df(df_pubon)
part_ops = res_futures(sub_pubon, req_ops_oeb)

part_ops2 = pd.merge(
missing_fr6.drop(columns=["address_complete_fr", "com_code", "ville", "dep_id", "dep_nom", "reg_nom"]),
part_ops,
on=["pn", "type_party", "sequence"], how="inner")

sub_ops2 = subset_df(part_ops2)

part_ops3 = res_futures(sub_ops2, adresses_ops_oeb)

addresses9 = addresses8.loc[~addresses8["key_appln_nr_person"].isin(part_ops3["key_appln_nr_person"])]
addresses9 = pd.concat([addresses9, part_ops3], ignore_index=True)

ad_siren = addresses9.loc[addresses9["siren"].notna(), ["docdb_family_id", "siren", "com_code",
"ville",
"dep_id",
"dep_nom",
"reg_nom"]].drop_duplicates().reset_index(
drop=True)

ad_siren = ad_siren.loc[ad_siren["com_code"].notna()].drop_duplicates().reset_index(drop=True)
ad_siren = ad_siren.loc[ad_siren["com_code"] != ""].drop_duplicates().reset_index(drop=True)

ad_siren_compte = ad_siren[["docdb_family_id", "siren", "com_code"]].groupby(["docdb_family_id", "siren"]).nunique(
dropna=False).reset_index().rename(
columns={"com_code": "compte"})

com_codeu = ad_siren_compte.loc[ad_siren_compte["compte"] == 1].reset_index(drop=True)

ad_sirenu = pd.merge(ad_siren, com_codeu, on=["docdb_family_id", "siren"], how="inner").drop(columns="compte")

addresses_siren_pm = addresses9.loc[
(addresses9["siren"].isin(com_codeu["siren"])) & (addresses9["type"] == "pm") & (
addresses9["com_code"].isna())].drop(columns=["com_code",
"ville",
"dep_id",
"dep_nom",
"reg_nom"]).drop_duplicates().reset_index(drop=True)

addresses_siren_pm = pd.merge(addresses_siren_pm, ad_sirenu, on=["docdb_family_id", "siren"], how="inner")

reste_addresses = addresses9.loc[~addresses9["key_appln_nr_person"].isin(addresses_siren_pm["key_appln_nr_person"])]

addresses10 = pd.concat([addresses_siren_pm, reste_addresses], ignore_index=True)

fam_name = addresses10.loc[addresses10["com_code"].notna(), ["docdb_family_id", "name_corrected",
"com_code",
"ville",
"dep_id",
"dep_nom",
"reg_nom"]].drop_duplicates().reset_index(
drop=True)

fam_name_compte = fam_name[["docdb_family_id", "name_corrected", "com_code"]].groupby(
["docdb_family_id", "name_corrected"]).nunique(dropna=False).reset_index().rename(
columns={"com_code": "compte"})

com_nomu = fam_name_compte.loc[fam_name_compte["compte"] == 1]

fam_nameu = pd.merge(fam_name, com_nomu, on=["docdb_family_id", "name_corrected"], how="inner")
fam_nameu = fam_nameu.drop(columns=["compte"])

addresses_fam_name = pd.merge(addresses10.loc[addresses10["com_code"].isna()].drop(columns=["com_code",
"ville",
"dep_id",
"dep_nom",
"reg_nom"]), fam_nameu,
on=["docdb_family_id", "name_corrected"],
how="inner").sort_values(
["docdb_family_id", "key_appln_nr_person"]).drop_duplicates().reset_index(drop=True)

reste_addresses2 = addresses10.loc[
~addresses10["key_appln_nr_person"].isin(addresses_fam_name["key_appln_nr_person"])]

addresses11 = pd.concat([addresses_fam_name, reste_addresses2], ignore_index=True)
addresses11 = addresses11.fillna("")

addresses11.to_csv("part_p08_address2.csv", index=False, sep="|", encoding="utf-8")
swift.upload_object('patstat', 'part_p08_address2.csv')

missing_fr7 = addresses11.loc[(addresses11["country_corrected"].isin(["FR", ""]))]
missing_fr8 = missing_fr7.loc[missing_fr7["com_code"] == ""]

pourcentage3 = round(len(missing_fr8) / len(missing_fr7) * 100, 2)
print(f"Le 3ème pourcentage d'adresses manquantes pour les Français est de {pourcentage3}.", flush=True)
missing_fr8.to_csv("missing_fr_address2.csv", index=False, sep="|", encoding="utf-8")
swift.upload_object('patstat', 'missing_fr_address2.csv')
missing_iloc = missing_fr6.iloc[0:100,:]

# df_pubon = pd.DataFrame(data={"pn": pubon})
sub_pubon = subset_df(missing_iloc)
addresses9 = res_futures(sub_pubon, req_ops_oeb)

addresses9.to_csv("test_p08_address2.csv", index=False, sep="|", encoding="utf-8")
swift.upload_object('patstat', 'test_p08_address2.csv')

# part_ops2 = pd.merge(
# missing_fr6.drop(columns=["address_complete_fr", "com_code", "ville", "dep_id", "dep_nom", "reg_nom"]),
# part_ops,
# on=["pn", "type_party", "sequence"], how="inner")
#
# sub_ops2 = subset_df(part_ops2)
#
# part_ops3 = res_futures(sub_ops2, adresses_ops_oeb)

# addresses9 = addresses8.loc[~addresses8["key_appln_nr_person"].isin(part_ops3["key_appln_nr_person"])]
# addresses9 = pd.concat([addresses9, part_ops3], ignore_index=True)
#
# ad_siren = addresses9.loc[addresses9["siren"].notna(), ["docdb_family_id", "siren", "com_code",
# "ville",
# "dep_id",
# "dep_nom",
# "reg_nom"]].drop_duplicates().reset_index(
# drop=True)
#
# ad_siren = ad_siren.loc[ad_siren["com_code"].notna()].drop_duplicates().reset_index(drop=True)
# ad_siren = ad_siren.loc[ad_siren["com_code"] != ""].drop_duplicates().reset_index(drop=True)
#
# ad_siren_compte = ad_siren[["docdb_family_id", "siren", "com_code"]].groupby(["docdb_family_id", "siren"]).nunique(
# dropna=False).reset_index().rename(
# columns={"com_code": "compte"})
#
# com_codeu = ad_siren_compte.loc[ad_siren_compte["compte"] == 1].reset_index(drop=True)
#
# ad_sirenu = pd.merge(ad_siren, com_codeu, on=["docdb_family_id", "siren"], how="inner").drop(columns="compte")
#
# addresses_siren_pm = addresses9.loc[
# (addresses9["siren"].isin(com_codeu["siren"])) & (addresses9["type"] == "pm") & (
# addresses9["com_code"].isna())].drop(columns=["com_code",
# "ville",
# "dep_id",
# "dep_nom",
# "reg_nom"]).drop_duplicates().reset_index(drop=True)
#
# addresses_siren_pm = pd.merge(addresses_siren_pm, ad_sirenu, on=["docdb_family_id", "siren"], how="inner")
#
# reste_addresses = addresses9.loc[~addresses9["key_appln_nr_person"].isin(addresses_siren_pm["key_appln_nr_person"])]
#
# addresses10 = pd.concat([addresses_siren_pm, reste_addresses], ignore_index=True)
#
# fam_name = addresses10.loc[addresses10["com_code"].notna(), ["docdb_family_id", "name_corrected",
# "com_code",
# "ville",
# "dep_id",
# "dep_nom",
# "reg_nom"]].drop_duplicates().reset_index(
# drop=True)
#
# fam_name_compte = fam_name[["docdb_family_id", "name_corrected", "com_code"]].groupby(
# ["docdb_family_id", "name_corrected"]).nunique(dropna=False).reset_index().rename(
# columns={"com_code": "compte"})
#
# com_nomu = fam_name_compte.loc[fam_name_compte["compte"] == 1]
#
# fam_nameu = pd.merge(fam_name, com_nomu, on=["docdb_family_id", "name_corrected"], how="inner")
# fam_nameu = fam_nameu.drop(columns=["compte"])
#
# addresses_fam_name = pd.merge(addresses10.loc[addresses10["com_code"].isna()].drop(columns=["com_code",
# "ville",
# "dep_id",
# "dep_nom",
# "reg_nom"]), fam_nameu,
# on=["docdb_family_id", "name_corrected"],
# how="inner").sort_values(
# ["docdb_family_id", "key_appln_nr_person"]).drop_duplicates().reset_index(drop=True)
#
# reste_addresses2 = addresses10.loc[
# ~addresses10["key_appln_nr_person"].isin(addresses_fam_name["key_appln_nr_person"])]
#
# addresses11 = pd.concat([addresses_fam_name, reste_addresses2], ignore_index=True)
# addresses11 = addresses11.fillna("")
#
# addresses11.to_csv("test_p08_address2.csv", index=False, sep="|", encoding="utf-8")
# swift.upload_object('patstat', 'test_p08_address2.csv')
#
# missing_fr7 = addresses11.loc[(addresses11["country_corrected"].isin(["FR", ""]))]
# missing_fr8 = missing_fr7.loc[missing_fr7["com_code"] == ""]
#
# pourcentage3 = round(len(missing_fr8) / len(missing_fr7) * 100, 2)
# print(f"Le 3ème pourcentage d'adresses manquantes pour les Français est de {pourcentage3}.", flush=True)
# missing_fr8.to_csv("missing_fr_address2.csv", index=False, sep="|", encoding="utf-8")
# swift.upload_object('patstat', 'missing_fr_address2.csv')

0 comments on commit 6de65ce

Please sign in to comment.