Skip to content

README.md

Latest commit

 

History

History
92 lines (63 loc) · 3.1 KB

README.md

File metadata and controls

92 lines (63 loc) · 3.1 KB
Raw

Dashboard — Joule Ledger

Pages

1. Executive Summary

Cumulative totals since 2011 (GJ saved, GHG avoided, $ saved) and progress against the current 5-year DSM Plan target. Slicers: year range, funding source.

Key visuals:

  • KPI cards: cumulative lifetime GJ, lifetime $ savings, current-year GJ beat/miss
  • Bar chart: annual GJ actuals with plan target overlay (2019–2024)
  • Donut: share of savings by category (Residential / Commercial / Industrial / Low-Income / Other)

2. Plan vs. Actual

Per-program comparison of filed targets vs. delivered actuals. The primary view for auditing program-level performance.

Key visuals:

  • Slope chart: target GJ vs. actual GJ, one line per program (filterable by year)
  • Table: program, target GJ, actual GJ, variance GJ, variance %, $/GJ cost-effectiveness
  • Highlight: programs that missed target appear in red; beats in green

Slicers: year, category, funding source, is_low_income.

3. Equity Lens

Dedicated view for low-income program tracking.

Key visuals:

  • Stacked bar: low-income vs. non-low-income share of annual spend (2019–2024)
  • Line chart: low-income participation count by year, split by funding source
  • KPI: low-income share of cumulative lifetime savings (~11.2%)
  • Table: program-level detail for is_low_income = 1 programs

4. Weather-Normalized Performance

Side-by-side view of raw actual savings and weather-normalized savings for electric programs.

Key visuals:

  • Clustered bar: actual_gj vs. actual_gj_wx_norm by year
  • Line: weather_factor by year (1.0 = normal year; >1 = colder than normal)
  • Annotation card: methodology note — HDD-ratio to 30-year ECCC baseline (see etl_design.md)

5. Methodology and Source Map

Every metric on every other page traced back to its source document and page number.

Content:

  • Table: metric → source document → page → extraction method → is_manually_entered count
  • Notes on unit conversions (GWh → GJ, restatement handling)
  • Link to docs/data_dictionary.md and docs/etl_design.md

Connecting to the warehouse

The model connects via ODBC to data/warehouse.db. To reconnect after moving the repo:

  1. Open dashboard.pbix in Power BI Desktop.
  2. Home → Transform data → Data source settings.
  3. Update the path to data/warehouse.db.
  4. Click Refresh.

The SQLite ODBC driver must be installed. Download from sqliteodbc.com (Windows) or via brew install sqliteodbc (macOS, for development only — Power BI requires Windows for full refresh).

DAX measures

All measures are defined in the model layer. Key measures:

[Variance GJ] = SUM(fact_actuals[actual_gj]) - SUM(fact_targets[target_gj])

[Variance %] =
    DIVIDE(
        SUM(fact_actuals[actual_gj]) - SUM(fact_targets[target_gj]),
        SUM(fact_targets[target_gj])
    )

[Actual $/GJ] =
    DIVIDE(
        SUM(fact_actuals[actual_spend_cad]),
        SUM(fact_actuals[actual_gj])
    )

[Low-Income Share of Spend %] =
    DIVIDE(
        CALCULATE(SUM(fact_actuals[actual_spend_cad]), dim_program[is_low_income] = 1),
        SUM(fact_actuals[actual_spend_cad])
    )