Uganda: Unregistered features
==============================

nonfood_expenditures
--------------------
Registered and aggregator fix applied (id_walk ordering), but
wave-level builds fail due to np.NaN removal in NumPy 2.0.

The error is in the upstream lsms package (lsms/tools.py:153),
specifically sum_food_expenditures_from_different_sources().

Fix: update the lsms package to replace np.NaN with np.nan.

plots
-----
Wave-level _/plots.py exists in all 8 waves. Needs:
1. Country-level aggregator: Uganda/_/plots.py
2. Makefile target for ../var/plots.parquet
3. data_scheme.yml entry: plots: !make
4. dvc.yaml foreach entry: uganda::::plots

education_expenses
------------------
Wave-level _/education_expenses.py exists in 7 waves (missing 2015-16). Needs:
1. Country-level aggregator: Uganda/_/education_expenses.py
2. Makefile target for ../var/education_expenses.parquet
3. data_scheme.yml entry: education_expenses: !make
4. dvc.yaml foreach entry: uganda::::education_expenses

Notes
-----
Both new aggregators should follow the assets.py pattern:
call id_walk() BEFORE set_index(['i', ...]).
