fix: exchange pat identifier with actual id, handle empty conditionda…

…te (#164)

* fix: exchange pat identifier with actual id, handle empty conditiondate
* update age at condition in case its empty to age at mid 2022

src/obds_fhir_to_opal/obds_fhir_to_opal.py

@@ -140,7 +140,11 @@ def calculate_age(birthdate):
 
 
 def calculate_age_at_conditiondate(birthdate, conditiondate):
-    age_at_conditiondate = conditiondate - birthdate
+    if conditiondate is None:
+        # todo: change this definitely
+        age_at_conditiondate = birthdate - datetime.date(2022, 6, 15)
+    else:
+        age_at_conditiondate = conditiondate - birthdate
     days_in_year = 365.2425
     age_at_conditiondate = int(age_at_conditiondate.days / days_in_year)
     return age_at_conditiondate
@@ -199,7 +203,8 @@ def add_age_at_condition_and_groups(df_pat_cond_joined):
     df_pat_cond_joined = df_pat_cond_joined.withColumn(
         "age_at_diagnosis",
         calculate_age_at_conditiondateUDF(
-            to_date(df_pat_cond_joined.birthDate), df_pat_cond_joined.conditiondate
+            to_date(df_pat_cond_joined.birthDate),
+            df_pat_cond_joined.conditiondate
         ),
     )
 
@@ -236,7 +241,8 @@ def encode_patients(ptl: PathlingContext, df_bundles: pyspark.sql.dataframe.Data
     return_yearUDF = udf(lambda x: return_year(x), StringType())
 
     patients = df_patients.selectExpr(
-        "id as pat_id", "gender", "birthDate", "deceasedBoolean", "deceasedDateTime"
+        "EXPLODE_OUTER(identifier.value) as pat_id", "gender", "birthDate",
+        "deceasedBoolean", "deceasedDateTime"
     )
 
     patients = patients.withColumns(