Add alternative-scenario toggle to gen_aec_data notebook

[alxsmith]

What

Add a high-level toggle (ALTERNATIVE_SCENARIO) to the config cell of gen_aec_data.qmd that switches the alternative HP adoption scenario between two modes:

• "hp_factor" (current behavior): BAU cumulative adoption scaled by heat_pump_adoption_factor
• "rie_high": uses Cumulative_High directly from the RIE adoption forecast Google Sheet

The toggle threads through all downstream code: adoption rate percentages (hp_pct_total), market-share lags (hp_hp_count_lag), the adoption chart, and ultimately the 8760 load curve generation via uniform_adoption.

Why

We need to compare BAU against the RIE High forecast (from the Google Sheet) as an alternative to our custom HP-factor scaling. Having a single toggle avoids duplicating the entire notebook and makes it easy to re-run with either assumption.

How

1. Add ALTERNATIVE_SCENARIO = "hp_factor" in the INPUT PARAMS cell, next to heat_pump_adoption_factor.
2. In the adoption-rates cell (~line 152), branch hp_pct_total on the toggle:
   • "hp_factor": heat_pump_adoption_factor * bau_hp_counts / total_num_homes (current logic)
   • "rie_high": get_adoption_forecast_num_homes(..., col_name="Cumulative_High") / total_num_homes
3. In the adoption chart cell (~line 161), branch hp_rates_cumulative:
   • "hp_factor": current growth-scaling logic
   • "rie_high": high_cumulative directly
4. In the hp_count_lag cell (~line 217), branch hp_hp_count_lag:
   • "hp_factor": bau_hp_count_lag * heat_pump_adoption_factor (current logic)
   • "rie_high": compute lags from Cumulative_High series
5. Everything downstream (hp_pct_upgrade_1, hp_scenario, df_hp_loads, GSheet output) consumes these variables unchanged.

Deliverables

• ALTERNATIVE_SCENARIO toggle added to config cell with "hp_factor" and "rie_high" options
• All three branching points implemented (adoption rates, chart data, market-share lags)
• Notebook runs end-to-end with both toggle values

[linear]

R2-102