Tighten source-backed target profile (#29)

Author: MaxGhenis

src/microplex_us/policyengine/target_profiles.py

@@ -761,6 +761,15 @@ def _target_cell_key(cell: PolicyEngineUSTargetCell) -> PolicyEngineUSTargetCell
         "The PolicyEngine cell is a pregnancy stock by state; live births are "
         "a flow and are not a defensible direct source fact for this target."
     ),
+    (
+        "person_count",
+        "state",
+        "adjusted_gross_income",
+        None,
+    ): (
+        "Loaded SOI state AGI sources provide return counts and AGI amounts, "
+        "not filer-person counts by AGI band."
+    ),
     (
         "alimony_expense",
         "national",
@@ -834,6 +843,71 @@ def _target_cell_key(cell: PolicyEngineUSTargetCell) -> PolicyEngineUSTargetCell
         "PolicyEngine rent is a household survey/model input; ACS rent tables "
         "do not provide a direct aggregate source fact for this exact variable."
     ),
+    (
+        "salt",
+        "national",
+        "salt",
+        None,
+    ): (
+        "SOI Table 2.1 itemized deduction sources cover itemizers; "
+        "PolicyEngine salt can be positive outside the itemizer domain."
+    ),
+    (
+        "tax_unit_count",
+        "national",
+        "salt",
+        None,
+    ): (
+        "SOI Table 2.1 publishes separate component counts, not the union "
+        "count of tax units with positive PolicyEngine salt."
+    ),
+    (
+        "tax_unit_count",
+        "national",
+        "salt,tax_unit_itemizes",
+        None,
+    ): (
+        "SOI Table 2.1 publishes separate component counts, not the union "
+        "count of itemizing tax units with positive PolicyEngine salt."
+    ),
+    (
+        "salt",
+        "state",
+        "salt",
+        None,
+    ): (
+        "Loaded state SOI sources do not provide an exact state-level "
+        "PolicyEngine salt amount; total state/local taxes also include "
+        "personal property taxes."
+    ),
+    (
+        "salt",
+        "state",
+        "salt,tax_unit_itemizes",
+        None,
+    ): (
+        "Loaded state SOI sources do not provide state-level itemizer salt "
+        "as income-or-sales tax plus real estate tax without personal "
+        "property taxes."
+    ),
+    (
+        "tax_unit_count",
+        "state",
+        "salt",
+        None,
+    ): (
+        "Loaded state SOI sources do not provide the union count of tax units "
+        "with positive PolicyEngine salt."
+    ),
+    (
+        "tax_unit_count",
+        "state",
+        "salt,tax_unit_itemizes",
+        None,
+    ): (
+        "Loaded state SOI sources do not provide the union count of itemizing "
+        "tax units with positive PolicyEngine salt."
+    ),
     (
         "spm_unit_capped_housing_subsidy",
         "national",

tests/policyengine/test_target_profiles.py

@@ -370,10 +370,10 @@ def test_source_backed_profile_excludes_only_documented_non_source_cells() -> No
     }
 
     assert len(broad_cells) == 189
-    assert len(exclusion_reasons) == 15
+    assert len(exclusion_reasons) == 23
     assert all(reason for reason in exclusion_reasons.values())
     assert set(exclusion_reasons) <= broad_cells
-    assert len(source_backed_cells) == 174
+    assert len(source_backed_cells) == 166
     assert source_backed_cells == broad_cells - set(exclusion_reasons)
     assert (
         "childcare_expenses",
@@ -387,6 +387,30 @@ def test_source_backed_profile_excludes_only_documented_non_source_cells() -> No
         "is_pregnant",
         None,
     ) not in source_backed_cells
+    assert (
+        "person_count",
+        "state",
+        "adjusted_gross_income",
+        None,
+    ) not in source_backed_cells
+    assert (
+        "salt",
+        "national",
+        "salt",
+        None,
+    ) not in source_backed_cells
+    assert (
+        "tax_unit_count",
+        "national",
+        "salt,tax_unit_itemizes",
+        None,
+    ) not in source_backed_cells
+    assert (
+        "salt",
+        "state",
+        "salt,tax_unit_itemizes",
+        None,
+    ) not in source_backed_cells
     assert (
         "employment_income",
         "national",
@@ -405,3 +429,9 @@ def test_source_backed_profile_excludes_only_documented_non_source_cells() -> No
         None,
         None,
     ) in source_backed_cells
+    assert (
+        "salt",
+        "national",
+        "salt,tax_unit_itemizes",
+        None,
+    ) in source_backed_cells