(
entity_df_with_features: dd.DataFrame,
df_to_join: dd.DataFrame,
join_keys: List[str],
)
| 1074 | |
| 1075 | |
| 1076 | def _merge( |
| 1077 | entity_df_with_features: dd.DataFrame, |
| 1078 | df_to_join: dd.DataFrame, |
| 1079 | join_keys: List[str], |
| 1080 | ) -> dd.DataFrame: |
| 1081 | # tmp join keys needed for cross join with null join table view |
| 1082 | tmp_join_keys = [] |
| 1083 | if not join_keys: |
| 1084 | entity_df_with_features["__tmp"] = 1 |
| 1085 | df_to_join["__tmp"] = 1 |
| 1086 | tmp_join_keys = ["__tmp"] |
| 1087 | |
| 1088 | # Get only data with requested entities |
| 1089 | df_to_join = dd.merge( |
| 1090 | entity_df_with_features, |
| 1091 | df_to_join, |
| 1092 | left_on=join_keys or tmp_join_keys, |
| 1093 | right_on=join_keys or tmp_join_keys, |
| 1094 | suffixes=("", "__"), |
| 1095 | how="left", |
| 1096 | ) |
| 1097 | |
| 1098 | if tmp_join_keys: |
| 1099 | df_to_join = df_to_join.drop(tmp_join_keys, axis=1).persist() |
| 1100 | else: |
| 1101 | df_to_join = df_to_join.persist() |
| 1102 | |
| 1103 | return df_to_join |
| 1104 | |
| 1105 | |
| 1106 | def _normalize_timestamp( |
no test coverage detected