ForecastResult class final Getting started
The signal projected past the end of the data.
For a local linear trend the variance of the forecast grows like the cube of the horizon, so the band widens like its three-halves power. This is expected: a slope that is free to wander soon becomes unknowable.
Properties
- hashCode → int
-
The hash code for this object.
no setterinherited
- length → int
-
Number of horizon points.
no setter
- mean → Float64List
-
Posterior mean of the signal at each horizon time.
final
- measurementVariance → double
-
The model's measurement variance, for predictiveInterval.
final
- runtimeType → Type
-
A representation of the runtime type of the object.
no setterinherited
- times → Float64List
-
The requested horizon, ascending.
final
- variance → Float64List
-
Its variance, excluding measurement noise.
final
Methods
-
credibleBand(
{double coverage = 0.95}) → Bands - credibleInterval at every horizon time, as two arrays to draw.
-
credibleInterval(
int i, {double coverage = 0.95}) → Band - Where the signal itself is heading.
-
noSuchMethod(
Invocation invocation) → dynamic -
Invoked when a nonexistent method or property is accessed.
inherited
-
predictiveBand(
{double coverage = 0.95}) → Bands - predictiveInterval at every horizon time, as two arrays to draw.
-
predictiveInterval(
int i, {double coverage = 0.95}) → Band - Where an actual future reading would fall: the same band widened by the measurement noise. This is the one to use for a question like "when will the scale read 75?", because the scale reading includes the noise.
-
toString(
) → String -
A string representation of this object.
inherited
Operators
-
operator ==(
Object other) → bool -
The equality operator.
inherited