Backtest Scores

Pool matched forecast rows across origins into per-variable and per-lead-time accuracy.

Cuts an already-scored comparison every way an accuracy review asks for.

It computes no forecast and fetches no actuals — it takes the comparison frame a backtest or a scoring emitted and re-aggregates it, which is why the same port id appears on both sides: the frame passes through unchanged beside the cuts.

At a glance

Module type	BacktestScores
Category	Intelligence
Default tier	durable
Runs on	reactive, durable
Starts a flow	no
Holds a body	no
Keeps state between runs	no
Available in	every application type

Parameters

metrics

string[] · optional · defaults to ["mape","wmape","mae","mse","rmse","bias","coverage"]

Which accuracy metrics to compute: mape, wmape, mae, mse, rmse, bias, coverage.

primaryMetric

string · optional · defaults to "wmape"

The one a KPI tile and the trend read; must be in metrics.

Inputs

Port	Kind	Required	Description
on	event	no	Implicit on every module that does not itself start a flow: wire an event here to make this module the start of what that event runs.
comparison	dataframe	yes	Forecast rows beside their actuals, from ForecastBacktest or ScoreForecast.

Outputs

Port	Kind	Connects to	Description
comparison	dataframe	dataframe, signal	The input frame, passed through so the cuts and their source stay wireable together.
byVariable	dataframe	dataframe, signal	Accuracy per forecast variable.
byHorizonVariable	dataframe	dataframe, signal	Accuracy per variable and lead time.
byHorizon	dataframe	dataframe, signal	Accuracy per lead time.
byOrigin	dataframe	dataframe, signal	Accuracy per forecast origin — which vintage of forecast was any good.
byEnvironment	dataframe	dataframe, signal	Accuracy per environment, for a panel twin.

---

← All application modules