def set_fk_provider_group(patient_df: DataFrame,
provider_group_df: DataFrame) -> DataFrame:
patient_df = transform_to_fk(patient_df, provider_group_df,
Constants.DATA_SOURCE_COLUMN,
Constants.DATA_SOURCE_COLUMN, "id",
"provider_group_id")
return patient_df
def set_fk_model(quality_assurance_df, model_df):
quality_assurance_df = quality_assurance_df.withColumnRenamed(
"model_id", "model_id_ref")
quality_assurance_df = transform_to_fk(quality_assurance_df, model_df,
"model_id_ref", "external_model_id",
"id", "model_id")
return quality_assurance_df
def set_fk_patient(sample_df: DataFrame, patient_df: DataFrame) -> DataFrame:
patient_snapshot_df = sample_df.withColumnRenamed("patient_id",
"patient_id_ref")
patient_snapshot_df = transform_to_fk(patient_snapshot_df, patient_df,
"patient_id_ref",
"external_patient_id", "id",
"patient_id")
return patient_snapshot_df
def set_fk_ethnicity(patient_df: DataFrame,
ethnicity_df: DataFrame) -> DataFrame:
patient_df = transform_to_fk(patient_df, ethnicity_df, "ethnicity", "name",
"id", "ethnicity_id")
return patient_df
def set_fk_diagnosis(raw_patient_df: DataFrame,
diagnosis_df: DataFrame) -> DataFrame:
patient_df = transform_to_fk(raw_patient_df, diagnosis_df,
"initial_diagnosis", "name", "id",
"initial_diagnosis_id")
return patient_df
def set_fk_sample_site(patient_sample_df: DataFrame,
tissue_df: DataFrame) -> DataFrame:
patient_sample_df = transform_to_fk(patient_sample_df, tissue_df,
"collection_site", "name", "id",
"collection_site_id")
return patient_sample_df
def set_fk_origin_tissue(patient_sample_df: DataFrame,
tissue_df: DataFrame) -> DataFrame:
patient_sample_df = transform_to_fk(patient_sample_df, tissue_df,
"primary_site", "name", "id",
"primary_site_id")
return patient_sample_df
def set_fk_diagnosis(patient_sample_df: DataFrame,
diagnosis_df: DataFrame) -> DataFrame:
patient_sample_df = transform_to_fk(patient_sample_df, diagnosis_df,
"diagnosis", "name", "id",
"diagnosis_id")
return patient_sample_df
def set_fk_model(patient_sample_df: DataFrame,
model_df: DataFrame) -> DataFrame:
patient_sample_df = transform_to_fk(patient_sample_df, model_df,
"model_name", "external_model_id",
"id", "model_id")
return patient_sample_df
def set_fk_tumour_type(patient_sample_df: DataFrame,
tumour_type_df: DataFrame) -> DataFrame:
patient_sample_df = transform_to_fk(patient_sample_df, tumour_type_df,
"tumour_type", "name", "id",
"tumour_type_id")
return patient_sample_df
def set_fk_provider_type(provider_group_df, provider_type_df):
provider_group_df = transform_to_fk(
provider_group_df, provider_type_df, "provider_type", "name", "id", "provider_type_id")
return provider_group_df
def set_fk_publication_group(model_df: DataFrame,
publication_group_df: DataFrame) -> DataFrame:
model_df = transform_to_fk(model_df, publication_group_df, "publications",
"pub_med_ids", "id", "publication_group_id")
return model_df
def set_fk_source_database(model_df: DataFrame,
source_database_df: DataFrame) -> DataFrame:
model_df = transform_to_fk(model_df, source_database_df, "database_url",
"database_url", "id", "source_database_id")
return model_df
def set_fk_contact_form(model_df: DataFrame,
contact_form_df: DataFrame) -> DataFrame:
model_df = transform_to_fk(model_df, contact_form_df, "form_url",
"form_url", "id", "contact_form_id")
return model_df