wip: debug syh dr data model

onefact · Apr 6, 2024 · 6dc81b7 · 6dc81b7
1 parent 6f595c6
commit 6dc81b7
Show file tree

Hide file tree

Showing 15 changed files with 185 additions and 128 deletions.
diff --git a/README.md b/README.md
@@ -100,21 +100,29 @@ cd healthcare_data
 python scripts/generate_syh_dr_data_models.py ~/data/syh_dr https://www.ahrq.gov/sites/default/files/wysiwyg/data/SyH-DR-Codebook.pdf
 ```
 
-2. Generate the synthetic healthcare data (takes ~5 minutes):
+2. Generate the synthetic healthcare data (takes ~2.5 minutes, with 8 threads on a Macbook):
 
 ```bash
-dbt run
+dbt run --threads 8
 ```
 
 3. Verify that you can query the data on the command line:
 
 ```bash
 
 
-## To build a specific data model:
+## To build a specific data model
 
 Use `--select` in dbt:
 
 ```bash
-syhdr_medicare_outpatient_2016
+dbt run --select "syhdr_medicare_outpatient_2016"
+```
+
+## To build a specific figure for visualization with Observable Framework
+
+Use `--select` in dbt to select models, e.g. in order to build all histograms:
+
+```bash
+dbt run --select "*histogram*"
 ```
diff --git a/healthcare_data/models/figures/insurance_plan_payment_histogram.sql b/healthcare_data/models/figures/insurance_plan_payment_histogram.sql
@@ -0,0 +1,39 @@
+{{ config(materialized='external', location=var('output_path') + '/' + this.name + '.parquet') }}
+
+WITH commercial_data AS (
+    SELECT
+        PLAN_PMT_AMT AS Payment,
+        COUNT(*) AS count,
+        'Commercial' AS Insurance
+    FROM read_parquet('/Users/me/data/syh_dr/syhdr_commercial_inpatient_2016.parquet')
+    GROUP BY PLAN_PMT_AMT
+),
+medicaid_data AS (
+    SELECT
+        PLAN_PMT_AMT AS Payment,
+        COUNT(*) AS count,
+        'Medicaid' AS Insurance
+    FROM read_parquet('/Users/me/data/syh_dr/syhdr_medicaid_inpatient_2016.parquet')
+    GROUP BY PLAN_PMT_AMT
+),
+medicare_data AS (
+    SELECT
+        PLAN_PMT_AMT AS Payment,
+        COUNT(*) AS count,
+        'Medicare' AS Insurance
+    FROM read_parquet('/Users/me/data/syh_dr/syhdr_medicare_inpatient_2016.parquet')
+    GROUP BY PLAN_PMT_AMT
+),
+combined_data AS (
+    SELECT * FROM commercial_data
+    UNION ALL
+    SELECT * FROM medicaid_data
+    UNION ALL
+    SELECT * FROM medicare_data
+)
+SELECT
+    Payment,
+    count,
+    Insurance
+FROM combined_data
+ORDER BY Insurance, Payment
diff --git a/healthcare_data/models/generated/with_types/syhdr_commercial_inpatient_2016.sql b/healthcare_data/models/generated/with_types/syhdr_commercial_inpatient_2016.sql
@@ -2,14 +2,14 @@
 {{ config(materialized='external', location=var('output_path') + '/' + this.name + '.parquet') }}
 
 SELECT
-    PERSON_ID::VARCHAR,
-    PERSON_WGHT::VARCHAR,
-    FACILITY_ID::VARCHAR,
-    CLM_CNTL_NUM::VARCHAR,
+    PERSON_ID::UBIGINT,
+    PERSON_WGHT::NUMERIC,
+    FACILITY_ID::UBIGINT,
+    CLM_CNTL_NUM::NUMERIC,
     AT_SPCLTY::VARCHAR,
-    SRVC_BEG_DATE::VARCHAR,
-    SRVC_END_DATE::VARCHAR,
-    LOS::VARCHAR,
+    SRVC_BEG_DATE::DATE,
+    SRVC_END_DATE::DATE,
+    LOS::UINTEGER,
     ADMSN_TYPE::VARCHAR,
     TOB_CD::VARCHAR,
     CLM_TYPE_CD::VARCHAR,
@@ -101,6 +101,6 @@ SELECT
     CPT_PRCDR_CD_33::VARCHAR,
     CPT_PRCDR_CD_34::VARCHAR,
     CPT_PRCDR_CD_35::VARCHAR,
-    PLAN_PMT_AMT::VARCHAR,
-    TOT_CHRG_AMT::VARCHAR
+    replace(replace(PLAN_PMT_AMT, '$', ''), ',', '')::FLOAT,
+    replace(replace(TOT_CHRG_AMT, '$', ''), ',', '')::FLOAT
 FROM read_csv('/Users/me/data/syh_dr/syhdr_commercial_inpatient_2016.CSV', header=True, null_padding=true, types={'CPT_PRCDR_CD_1': 'VARCHAR', 'CPT_PRCDR_CD_2': 'VARCHAR', 'CPT_PRCDR_CD_3': 'VARCHAR', 'CPT_PRCDR_CD_4': 'VARCHAR', 'CPT_PRCDR_CD_5': 'VARCHAR', 'CPT_PRCDR_CD_6': 'VARCHAR', 'CPT_PRCDR_CD_7': 'VARCHAR', 'CPT_PRCDR_CD_8': 'VARCHAR', 'CPT_PRCDR_CD_9': 'VARCHAR', 'CPT_PRCDR_CD_10': 'VARCHAR', 'CPT_PRCDR_CD_11': 'VARCHAR', 'CPT_PRCDR_CD_12': 'VARCHAR', 'CPT_PRCDR_CD_13': 'VARCHAR', 'CPT_PRCDR_CD_14': 'VARCHAR', 'CPT_PRCDR_CD_15': 'VARCHAR', 'CPT_PRCDR_CD_16': 'VARCHAR', 'CPT_PRCDR_CD_17': 'VARCHAR', 'CPT_PRCDR_CD_18': 'VARCHAR', 'CPT_PRCDR_CD_19': 'VARCHAR', 'CPT_PRCDR_CD_20': 'VARCHAR', 'CPT_PRCDR_CD_21': 'VARCHAR', 'CPT_PRCDR_CD_22': 'VARCHAR', 'CPT_PRCDR_CD_23': 'VARCHAR', 'CPT_PRCDR_CD_24': 'VARCHAR', 'CPT_PRCDR_CD_25': 'VARCHAR', 'CPT_PRCDR_CD_26': 'VARCHAR', 'CPT_PRCDR_CD_27': 'VARCHAR', 'CPT_PRCDR_CD_28': 'VARCHAR', 'CPT_PRCDR_CD_29': 'VARCHAR', 'CPT_PRCDR_CD_30': 'VARCHAR', 'CPT_PRCDR_CD_31': 'VARCHAR', 'CPT_PRCDR_CD_32': 'VARCHAR', 'CPT_PRCDR_CD_33': 'VARCHAR', 'CPT_PRCDR_CD_34': 'VARCHAR', 'CPT_PRCDR_CD_35': 'VARCHAR'}, ignore_errors=true)
diff --git a/healthcare_data/models/generated/with_types/syhdr_commercial_outpatient_2016.sql b/healthcare_data/models/generated/with_types/syhdr_commercial_outpatient_2016.sql
@@ -2,14 +2,14 @@
 {{ config(materialized='external', location=var('output_path') + '/' + this.name + '.parquet') }}
 
 SELECT
-    PERSON_ID::VARCHAR,
-    PERSON_WGHT::VARCHAR,
-    FACILITY_ID::VARCHAR,
-    CLM_CNTL_NUM::VARCHAR,
+    PERSON_ID::UBIGINT,
+    PERSON_WGHT::NUMERIC,
+    FACILITY_ID::UBIGINT,
+    CLM_CNTL_NUM::NUMERIC,
     AT_SPCLTY::VARCHAR,
-    SRVC_BEG_DATE::VARCHAR,
-    SRVC_END_DATE::VARCHAR,
-    LOS::VARCHAR,
+    SRVC_BEG_DATE::DATE,
+    SRVC_END_DATE::DATE,
+    LOS::UINTEGER,
     ADMSN_TYPE::VARCHAR,
     TOB_CD::VARCHAR,
     CLM_TYPE_CD::VARCHAR,
@@ -101,6 +101,6 @@ SELECT
     CPT_PRCDR_CD_33::VARCHAR,
     CPT_PRCDR_CD_34::VARCHAR,
     CPT_PRCDR_CD_35::VARCHAR,
-    PLAN_PMT_AMT::VARCHAR,
-    TOT_CHRG_AMT::VARCHAR
+    replace(replace(PLAN_PMT_AMT, '$', ''), ',', '')::FLOAT,
+    replace(replace(TOT_CHRG_AMT, '$', ''), ',', '')::FLOAT
 FROM read_csv('/Users/me/data/syh_dr/syhdr_commercial_outpatient_2016.CSV', header=True, null_padding=true, types={'CPT_PRCDR_CD_1': 'VARCHAR', 'CPT_PRCDR_CD_2': 'VARCHAR', 'CPT_PRCDR_CD_3': 'VARCHAR', 'CPT_PRCDR_CD_4': 'VARCHAR', 'CPT_PRCDR_CD_5': 'VARCHAR', 'CPT_PRCDR_CD_6': 'VARCHAR', 'CPT_PRCDR_CD_7': 'VARCHAR', 'CPT_PRCDR_CD_8': 'VARCHAR', 'CPT_PRCDR_CD_9': 'VARCHAR', 'CPT_PRCDR_CD_10': 'VARCHAR', 'CPT_PRCDR_CD_11': 'VARCHAR', 'CPT_PRCDR_CD_12': 'VARCHAR', 'CPT_PRCDR_CD_13': 'VARCHAR', 'CPT_PRCDR_CD_14': 'VARCHAR', 'CPT_PRCDR_CD_15': 'VARCHAR', 'CPT_PRCDR_CD_16': 'VARCHAR', 'CPT_PRCDR_CD_17': 'VARCHAR', 'CPT_PRCDR_CD_18': 'VARCHAR', 'CPT_PRCDR_CD_19': 'VARCHAR', 'CPT_PRCDR_CD_20': 'VARCHAR', 'CPT_PRCDR_CD_21': 'VARCHAR', 'CPT_PRCDR_CD_22': 'VARCHAR', 'CPT_PRCDR_CD_23': 'VARCHAR', 'CPT_PRCDR_CD_24': 'VARCHAR', 'CPT_PRCDR_CD_25': 'VARCHAR', 'CPT_PRCDR_CD_26': 'VARCHAR', 'CPT_PRCDR_CD_27': 'VARCHAR', 'CPT_PRCDR_CD_28': 'VARCHAR', 'CPT_PRCDR_CD_29': 'VARCHAR', 'CPT_PRCDR_CD_30': 'VARCHAR', 'CPT_PRCDR_CD_31': 'VARCHAR', 'CPT_PRCDR_CD_32': 'VARCHAR', 'CPT_PRCDR_CD_33': 'VARCHAR', 'CPT_PRCDR_CD_34': 'VARCHAR', 'CPT_PRCDR_CD_35': 'VARCHAR'}, ignore_errors=true)
diff --git a/healthcare_data/models/generated/with_types/syhdr_commercial_person_2016.sql b/healthcare_data/models/generated/with_types/syhdr_commercial_person_2016.sql
@@ -2,15 +2,15 @@
 {{ config(materialized='external', location=var('output_path') + '/' + this.name + '.parquet') }}
 
 SELECT
-    PERSON_ID::VARCHAR,
-    PERSON_WGHT::VARCHAR,
-    AGE_LOW::VARCHAR,
-    AGE_HIGH::VARCHAR,
+    PERSON_ID::UBIGINT,
+    PERSON_WGHT::NUMERIC,
+    AGE_LOW::NUMERIC,
+    AGE_HIGH::NUMERIC,
     SEX_IDENT_CD::VARCHAR,
     STATE_CD::VARCHAR,
     COUNTY_FIPS_CD::VARCHAR,
     ZIP_CD::VARCHAR,
-    PHRMCY_CVRG_1::VARCHAR,
+    PHRMCY_CVRG_1::NUMERIC,
     PHRMCY_CVRG_2::VARCHAR,
     PHRMCY_CVRG_3::VARCHAR,
     PHRMCY_CVRG_4::VARCHAR,
@@ -21,8 +21,8 @@ SELECT
     PHRMCY_CVRG_9::VARCHAR,
     PHRMCY_CVRG_10::VARCHAR,
     PHRMCY_CVRG_11::VARCHAR,
-    PHRMCY_CVRG_12::VARCHAR,
-    CMRCL_INSRC_1::VARCHAR,
+    PHRMCY_CVRG_12::NUMERIC,
+    CMRCL_INSRC_1::NUMERIC,
     CMRCL_INSRC_2::VARCHAR,
     CMRCL_INSRC_3::VARCHAR,
     CMRCL_INSRC_4::VARCHAR,
@@ -33,5 +33,5 @@ SELECT
     CMRCL_INSRC_9::VARCHAR,
     CMRCL_INSRC_10::VARCHAR,
     CMRCL_INSRC_11::VARCHAR,
-    CMRCL_INSRC_12::VARCHAR
+    CMRCL_INSRC_12::NUMERIC
 FROM read_csv('/Users/me/data/syh_dr/syhdr_commercial_person_2016.CSV', header=True, null_padding=true)
diff --git a/healthcare_data/models/generated/with_types/syhdr_commercial_pharmacy_2016.sql b/healthcare_data/models/generated/with_types/syhdr_commercial_pharmacy_2016.sql
@@ -2,14 +2,14 @@
 {{ config(materialized='external', location=var('output_path') + '/' + this.name + '.parquet') }}
 
 SELECT
-    PERSON_ID::VARCHAR,
-    PERSON_WGHT::VARCHAR,
-    PHMCY_CLM_NUM::VARCHAR,
-    CLM_CNTL_NUM::VARCHAR,
+    PERSON_ID::UBIGINT,
+    PERSON_WGHT::NUMERIC,
+    PHMCY_CLM_NUM::NUMERIC,
+    CLM_CNTL_NUM::NUMERIC,
     LINE_NBR::VARCHAR,
-    FILL_DT::VARCHAR,
+    FILL_DT::DATE,
     SYNTHETIC_DRUG_ID::VARCHAR,
     GENERIC_DRUG_NAME::VARCHAR,
-    PLAN_PMT_AMT::VARCHAR,
-    TOT_CHRG_AMT::VARCHAR
+    replace(replace(PLAN_PMT_AMT, '$', ''), ',', '')::FLOAT,
+    replace(replace(TOT_CHRG_AMT, '$', ''), ',', '')::FLOAT
 FROM read_csv('/Users/me/data/syh_dr/syhdr_commercial_pharmacy_2016.CSV', header=True, null_padding=true)
diff --git a/healthcare_data/models/generated/with_types/syhdr_medicaid_inpatient_2016.sql b/healthcare_data/models/generated/with_types/syhdr_medicaid_inpatient_2016.sql
@@ -2,14 +2,14 @@
 {{ config(materialized='external', location=var('output_path') + '/' + this.name + '.parquet') }}
 
 SELECT
-    PERSON_ID::VARCHAR,
-    PERSON_WGHT::VARCHAR,
-    FACILITY_ID::VARCHAR,
-    CLM_CNTL_NUM::VARCHAR,
+    PERSON_ID::UBIGINT,
+    PERSON_WGHT::NUMERIC,
+    FACILITY_ID::UBIGINT,
+    CLM_CNTL_NUM::NUMERIC,
     AT_SPCLTY::VARCHAR,
-    SRVC_BEG_DATE::VARCHAR,
-    SRVC_END_DATE::VARCHAR,
-    LOS::VARCHAR,
+    SRVC_BEG_DATE::DATE,
+    SRVC_END_DATE::DATE,
+    LOS::UINTEGER,
     ADMSN_TYPE::VARCHAR,
     TOB_CD::VARCHAR,
     CLM_TYPE_CD::VARCHAR,
@@ -101,6 +101,6 @@ SELECT
     CPT_PRCDR_CD_33::VARCHAR,
     CPT_PRCDR_CD_34::VARCHAR,
     CPT_PRCDR_CD_35::VARCHAR,
-    PLAN_PMT_AMT::VARCHAR,
-    TOT_CHRG_AMT::VARCHAR
+    replace(replace(PLAN_PMT_AMT, '$', ''), ',', '')::FLOAT,
+    replace(replace(TOT_CHRG_AMT, '$', ''), ',', '')::FLOAT
 FROM read_csv('/Users/me/data/syh_dr/syhdr_medicaid_inpatient_2016.CSV', header=True, null_padding=true, types={'CPT_PRCDR_CD_1': 'VARCHAR', 'CPT_PRCDR_CD_2': 'VARCHAR', 'CPT_PRCDR_CD_3': 'VARCHAR', 'CPT_PRCDR_CD_4': 'VARCHAR', 'CPT_PRCDR_CD_5': 'VARCHAR', 'CPT_PRCDR_CD_6': 'VARCHAR', 'CPT_PRCDR_CD_7': 'VARCHAR', 'CPT_PRCDR_CD_8': 'VARCHAR', 'CPT_PRCDR_CD_9': 'VARCHAR', 'CPT_PRCDR_CD_10': 'VARCHAR', 'CPT_PRCDR_CD_11': 'VARCHAR', 'CPT_PRCDR_CD_12': 'VARCHAR', 'CPT_PRCDR_CD_13': 'VARCHAR', 'CPT_PRCDR_CD_14': 'VARCHAR', 'CPT_PRCDR_CD_15': 'VARCHAR', 'CPT_PRCDR_CD_16': 'VARCHAR', 'CPT_PRCDR_CD_17': 'VARCHAR', 'CPT_PRCDR_CD_18': 'VARCHAR', 'CPT_PRCDR_CD_19': 'VARCHAR', 'CPT_PRCDR_CD_20': 'VARCHAR', 'CPT_PRCDR_CD_21': 'VARCHAR', 'CPT_PRCDR_CD_22': 'VARCHAR', 'CPT_PRCDR_CD_23': 'VARCHAR', 'CPT_PRCDR_CD_24': 'VARCHAR', 'CPT_PRCDR_CD_25': 'VARCHAR', 'CPT_PRCDR_CD_26': 'VARCHAR', 'CPT_PRCDR_CD_27': 'VARCHAR', 'CPT_PRCDR_CD_28': 'VARCHAR', 'CPT_PRCDR_CD_29': 'VARCHAR', 'CPT_PRCDR_CD_30': 'VARCHAR', 'CPT_PRCDR_CD_31': 'VARCHAR', 'CPT_PRCDR_CD_32': 'VARCHAR', 'CPT_PRCDR_CD_33': 'VARCHAR', 'CPT_PRCDR_CD_34': 'VARCHAR', 'CPT_PRCDR_CD_35': 'VARCHAR'}, ignore_errors=true)
diff --git a/healthcare_data/models/generated/with_types/syhdr_medicaid_outpatient_2016.sql b/healthcare_data/models/generated/with_types/syhdr_medicaid_outpatient_2016.sql
@@ -2,14 +2,14 @@
 {{ config(materialized='external', location=var('output_path') + '/' + this.name + '.parquet') }}
 
 SELECT
-    PERSON_ID::VARCHAR,
-    PERSON_WGHT::VARCHAR,
-    FACILITY_ID::VARCHAR,
-    CLM_CNTL_NUM::VARCHAR,
+    PERSON_ID::UBIGINT,
+    PERSON_WGHT::NUMERIC,
+    FACILITY_ID::UBIGINT,
+    CLM_CNTL_NUM::NUMERIC,
     AT_SPCLTY::VARCHAR,
-    SRVC_BEG_DATE::VARCHAR,
-    SRVC_END_DATE::VARCHAR,
-    LOS::VARCHAR,
+    SRVC_BEG_DATE::DATE,
+    SRVC_END_DATE::DATE,
+    LOS::UINTEGER,
     ADMSN_TYPE::VARCHAR,
     TOB_CD::VARCHAR,
     CLM_TYPE_CD::VARCHAR,
@@ -101,6 +101,6 @@ SELECT
     CPT_PRCDR_CD_33::VARCHAR,
     CPT_PRCDR_CD_34::VARCHAR,
     CPT_PRCDR_CD_35::VARCHAR,
-    PLAN_PMT_AMT::VARCHAR,
-    TOT_CHRG_AMT::VARCHAR
+    replace(replace(PLAN_PMT_AMT, '$', ''), ',', '')::FLOAT,
+    replace(replace(TOT_CHRG_AMT, '$', ''), ',', '')::FLOAT
 FROM read_csv('/Users/me/data/syh_dr/syhdr_medicaid_outpatient_2016.CSV', header=True, null_padding=true, types={'CPT_PRCDR_CD_1': 'VARCHAR', 'CPT_PRCDR_CD_2': 'VARCHAR', 'CPT_PRCDR_CD_3': 'VARCHAR', 'CPT_PRCDR_CD_4': 'VARCHAR', 'CPT_PRCDR_CD_5': 'VARCHAR', 'CPT_PRCDR_CD_6': 'VARCHAR', 'CPT_PRCDR_CD_7': 'VARCHAR', 'CPT_PRCDR_CD_8': 'VARCHAR', 'CPT_PRCDR_CD_9': 'VARCHAR', 'CPT_PRCDR_CD_10': 'VARCHAR', 'CPT_PRCDR_CD_11': 'VARCHAR', 'CPT_PRCDR_CD_12': 'VARCHAR', 'CPT_PRCDR_CD_13': 'VARCHAR', 'CPT_PRCDR_CD_14': 'VARCHAR', 'CPT_PRCDR_CD_15': 'VARCHAR', 'CPT_PRCDR_CD_16': 'VARCHAR', 'CPT_PRCDR_CD_17': 'VARCHAR', 'CPT_PRCDR_CD_18': 'VARCHAR', 'CPT_PRCDR_CD_19': 'VARCHAR', 'CPT_PRCDR_CD_20': 'VARCHAR', 'CPT_PRCDR_CD_21': 'VARCHAR', 'CPT_PRCDR_CD_22': 'VARCHAR', 'CPT_PRCDR_CD_23': 'VARCHAR', 'CPT_PRCDR_CD_24': 'VARCHAR', 'CPT_PRCDR_CD_25': 'VARCHAR', 'CPT_PRCDR_CD_26': 'VARCHAR', 'CPT_PRCDR_CD_27': 'VARCHAR', 'CPT_PRCDR_CD_28': 'VARCHAR', 'CPT_PRCDR_CD_29': 'VARCHAR', 'CPT_PRCDR_CD_30': 'VARCHAR', 'CPT_PRCDR_CD_31': 'VARCHAR', 'CPT_PRCDR_CD_32': 'VARCHAR', 'CPT_PRCDR_CD_33': 'VARCHAR', 'CPT_PRCDR_CD_34': 'VARCHAR', 'CPT_PRCDR_CD_35': 'VARCHAR'}, ignore_errors=true)
diff --git a/healthcare_data/models/generated/with_types/syhdr_medicaid_person_2016.sql b/healthcare_data/models/generated/with_types/syhdr_medicaid_person_2016.sql
@@ -2,19 +2,19 @@
 {{ config(materialized='external', location=var('output_path') + '/' + this.name + '.parquet') }}
 
 SELECT
-    PERSON_ID::VARCHAR,
-    MCAID_BENE_ID::VARCHAR,
-    PERSON_WGHT::VARCHAR,
-    AGE_LOW::VARCHAR,
-    AGE_HIGH::VARCHAR,
+    PERSON_ID::UBIGINT,
+    MCAID_BENE_ID::UBIGINT,
+    PERSON_WGHT::NUMERIC,
+    AGE_LOW::NUMERIC,
+    AGE_HIGH::NUMERIC,
     SEX_IDENT_CD::VARCHAR,
     RACE_CD::VARCHAR,
     MCAID_SBMTTG_ST_CD::VARCHAR,
     STATE_CD::VARCHAR,
     COUNTY_FIPS_CD::VARCHAR,
     ZIP_CD::VARCHAR,
     RSN_ENRLMT_CD::VARCHAR,
-    MDCD_ENRLMT_1::VARCHAR,
+    MDCD_ENRLMT_1::NUMERIC,
     MDCD_ENRLMT_2::VARCHAR,
     MDCD_ENRLMT_3::VARCHAR,
     MDCD_ENRLMT_4::VARCHAR,
@@ -25,8 +25,8 @@ SELECT
     MDCD_ENRLMT_9::VARCHAR,
     MDCD_ENRLMT_10::VARCHAR,
     MDCD_ENRLMT_11::VARCHAR,
-    MDCD_ENRLMT_12::VARCHAR,
-    MDCD_MCO_ENRLMT_1::VARCHAR,
+    MDCD_ENRLMT_12::NUMERIC,
+    MDCD_MCO_ENRLMT_1::NUMERIC,
     MDCD_MCO_ENRLMT_2::VARCHAR,
     MDCD_MCO_ENRLMT_3::VARCHAR,
     MDCD_MCO_ENRLMT_4::VARCHAR,
@@ -37,10 +37,10 @@ SELECT
     MDCD_MCO_ENRLMT_9::VARCHAR,
     MDCD_MCO_ENRLMT_10::VARCHAR,
     MDCD_MCO_ENRLMT_11::VARCHAR,
-    MDCD_MCO_ENRLMT_12::VARCHAR,
-    MDCD_CHIP_ENRLMT::VARCHAR,
+    MDCD_MCO_ENRLMT_12::NUMERIC,
+    MDCD_CHIP_ENRLMT::NUMERIC,
     RSTRCTD_BNFTS_IND::VARCHAR,
-    DUAL_ELGBL_1::VARCHAR,
+    DUAL_ELGBL_1::NUMERIC,
     DUAL_ELGBL_2::VARCHAR,
     DUAL_ELGBL_3::VARCHAR,
     DUAL_ELGBL_4::VARCHAR,
@@ -51,5 +51,5 @@ SELECT
     DUAL_ELGBL_9::VARCHAR,
     DUAL_ELGBL_10::VARCHAR,
     DUAL_ELGBL_11::VARCHAR,
-    DUAL_ELGBL_12::VARCHAR
+    DUAL_ELGBL_12::NUMERIC
 FROM read_csv('/Users/me/data/syh_dr/syhdr_medicaid_person_2016.CSV', header=True, null_padding=true)
diff --git a/healthcare_data/models/generated/with_types/syhdr_medicaid_pharmacy_2016.sql b/healthcare_data/models/generated/with_types/syhdr_medicaid_pharmacy_2016.sql
@@ -2,14 +2,14 @@
 {{ config(materialized='external', location=var('output_path') + '/' + this.name + '.parquet') }}
 
 SELECT
-    PERSON_ID::VARCHAR,
-    PERSON_WGHT::VARCHAR,
-    PHMCY_CLM_NUM::VARCHAR,
-    CLM_CNTL_NUM::VARCHAR,
+    PERSON_ID::UBIGINT,
+    PERSON_WGHT::NUMERIC,
+    PHMCY_CLM_NUM::NUMERIC,
+    CLM_CNTL_NUM::NUMERIC,
     LINE_NBR::VARCHAR,
-    FILL_DT::VARCHAR,
+    FILL_DT::DATE,
     SYNTHETIC_DRUG_ID::VARCHAR,
     GENERIC_DRUG_NAME::VARCHAR,
-    PLAN_PMT_AMT::VARCHAR,
-    TOT_CHRG_AMT::VARCHAR
+    replace(replace(PLAN_PMT_AMT, '$', ''), ',', '')::FLOAT,
+    replace(replace(TOT_CHRG_AMT, '$', ''), ',', '')::FLOAT
 FROM read_csv('/Users/me/data/syh_dr/syhdr_medicaid_pharmacy_2016.CSV', header=True, null_padding=true)
diff --git a/healthcare_data/models/generated/with_types/syhdr_medicare_inpatient_2016.sql b/healthcare_data/models/generated/with_types/syhdr_medicare_inpatient_2016.sql
@@ -2,14 +2,14 @@
 {{ config(materialized='external', location=var('output_path') + '/' + this.name + '.parquet') }}
 
 SELECT
-    PERSON_ID::VARCHAR,
-    PERSON_WGHT::VARCHAR,
-    FACILITY_ID::VARCHAR,
-    CLM_CNTL_NUM::VARCHAR,
+    PERSON_ID::UBIGINT,
+    PERSON_WGHT::NUMERIC,
+    FACILITY_ID::UBIGINT,
+    CLM_CNTL_NUM::NUMERIC,
     AT_SPCLTY::VARCHAR,
-    SRVC_BEG_DATE::VARCHAR,
-    SRVC_END_DATE::VARCHAR,
-    LOS::VARCHAR,
+    SRVC_BEG_DATE::DATE,
+    SRVC_END_DATE::DATE,
+    LOS::UINTEGER,
     ADMSN_TYPE::VARCHAR,
     TOB_CD::VARCHAR,
     CLM_TYPE_CD::VARCHAR,
@@ -101,6 +101,6 @@ SELECT
     CPT_PRCDR_CD_33::VARCHAR,
     CPT_PRCDR_CD_34::VARCHAR,
     CPT_PRCDR_CD_35::VARCHAR,
-    PLAN_PMT_AMT::VARCHAR,
-    TOT_CHRG_AMT::VARCHAR
+    replace(replace(PLAN_PMT_AMT, '$', ''), ',', '')::FLOAT,
+    replace(replace(TOT_CHRG_AMT, '$', ''), ',', '')::FLOAT
 FROM read_csv('/Users/me/data/syh_dr/syhdr_medicare_inpatient_2016.CSV', header=True, null_padding=true, types={'CPT_PRCDR_CD_1': 'VARCHAR', 'CPT_PRCDR_CD_2': 'VARCHAR', 'CPT_PRCDR_CD_3': 'VARCHAR', 'CPT_PRCDR_CD_4': 'VARCHAR', 'CPT_PRCDR_CD_5': 'VARCHAR', 'CPT_PRCDR_CD_6': 'VARCHAR', 'CPT_PRCDR_CD_7': 'VARCHAR', 'CPT_PRCDR_CD_8': 'VARCHAR', 'CPT_PRCDR_CD_9': 'VARCHAR', 'CPT_PRCDR_CD_10': 'VARCHAR', 'CPT_PRCDR_CD_11': 'VARCHAR', 'CPT_PRCDR_CD_12': 'VARCHAR', 'CPT_PRCDR_CD_13': 'VARCHAR', 'CPT_PRCDR_CD_14': 'VARCHAR', 'CPT_PRCDR_CD_15': 'VARCHAR', 'CPT_PRCDR_CD_16': 'VARCHAR', 'CPT_PRCDR_CD_17': 'VARCHAR', 'CPT_PRCDR_CD_18': 'VARCHAR', 'CPT_PRCDR_CD_19': 'VARCHAR', 'CPT_PRCDR_CD_20': 'VARCHAR', 'CPT_PRCDR_CD_21': 'VARCHAR', 'CPT_PRCDR_CD_22': 'VARCHAR', 'CPT_PRCDR_CD_23': 'VARCHAR', 'CPT_PRCDR_CD_24': 'VARCHAR', 'CPT_PRCDR_CD_25': 'VARCHAR', 'CPT_PRCDR_CD_26': 'VARCHAR', 'CPT_PRCDR_CD_27': 'VARCHAR', 'CPT_PRCDR_CD_28': 'VARCHAR', 'CPT_PRCDR_CD_29': 'VARCHAR', 'CPT_PRCDR_CD_30': 'VARCHAR', 'CPT_PRCDR_CD_31': 'VARCHAR', 'CPT_PRCDR_CD_32': 'VARCHAR', 'CPT_PRCDR_CD_33': 'VARCHAR', 'CPT_PRCDR_CD_34': 'VARCHAR', 'CPT_PRCDR_CD_35': 'VARCHAR'}, ignore_errors=true)