Small ECPS is too large (#308)

* Small CPS is too large
Fixes #307

* Rename to ECPS

Author: nikhilwoodruff

Makefile

@@ -42,7 +42,7 @@ data:
 	python policyengine_us_data/datasets/puf/puf.py
 	python policyengine_us_data/datasets/cps/extended_cps.py
 	python policyengine_us_data/datasets/cps/enhanced_cps.py
-	python policyengine_us_data/datasets/cps/small_cps.py
+	python policyengine_us_data/datasets/cps/small_enhanced_cps.py
 
 clean:
 	rm -f policyengine_us_data/storage/*.h5

changelog_entry.yaml

@@ -0,0 +1,4 @@
+- bump: patch
+  changes:
+    fixed:
+    - Small CPS is now 1000 households.

…engine_us_data/datasets/cps/small_cps.py …_data/datasets/cps/small_enhanced_cps.pypolicyengine_us_data/datasets/cps/small_cps.py renamed to policyengine_us_data/datasets/cps/small_enhanced_cps.py policyengine_us_data/datasets/cps/small_cps.py renamed to policyengine_us_data/datasets/cps/small_enhanced_cps.py

@@ -1,7 +1,7 @@
 import numpy as np
 
 
-def create_small_cps():
+def create_small_ecps():
     from policyengine_us import Microsimulation
     from policyengine_us_data.datasets import EnhancedCPS_2024
     from policyengine_us_data.storage import STORAGE_FOLDER
@@ -10,6 +10,7 @@ def create_small_cps():
     simulation = Microsimulation(
         dataset=EnhancedCPS_2024,
     )
+    simulation.subsample(1_000)
 
     data = {}
     for variable in simulation.tax_benefit_system.variables:
@@ -29,13 +30,13 @@ def create_small_cps():
 
     import h5py
 
-    with h5py.File(STORAGE_FOLDER / "small_cps_2024.h5", "w") as f:
+    with h5py.File(STORAGE_FOLDER / "small_enhanced_cps_2024.h5", "w") as f:
         for variable, periods in data.items():
             grp = f.create_group(variable)
             for period, values in periods.items():
                 grp.create_dataset(str(period), data=values)
 
 
 if __name__ == "__main__":
-    create_small_cps()
+    create_small_ecps()
     print("Small CPS dataset created successfully.")

policyengine_us_data/storage/upload_completed_datasets.py

@@ -14,7 +14,7 @@ def upload_datasets():
         EnhancedCPS_2024.file_path,
         Pooled_3_Year_CPS_2023.file_path,
         CPS_2023.file_path,
-        STORAGE_FOLDER / "small_cps_2024.h5",
+        STORAGE_FOLDER / "small_enhanced_cps_2024.h5",
     ]
 
     for file_path in dataset_files:

…ta/tests/test_datasets/test_small_cps.py …test_datasets/test_small_enhanced_cps.pypolicyengine_us_data/tests/test_datasets/test_small_cps.py renamed to policyengine_us_data/tests/test_datasets/test_small_enhanced_cps.py policyengine_us_data/tests/test_datasets/test_small_cps.py renamed to policyengine_us_data/tests/test_datasets/test_small_enhanced_cps.py

@@ -3,14 +3,14 @@
 
 
 @pytest.mark.parametrize("year", [2024])
-def test_small_cps_loads(year: int):
+def test_small_ecps_loads(year: int):
     from policyengine_core.data import Dataset
     from policyengine_us_data.storage import STORAGE_FOLDER
     from policyengine_us import Microsimulation
 
     sim = Microsimulation(
         dataset=Dataset.from_file(
-            STORAGE_FOLDER / f"small_cps_{year}.h5",
+            STORAGE_FOLDER / f"small_enhanced_cps_{year}.h5",
         )
     )