TimeSeriesCatalog.ForecastBySsa Method
Definition
Important
Some information relates to prerelease product that may be substantially modified before it’s released. Microsoft makes no warranties, express or implied, with respect to the information provided here.
Singular Spectrum Analysis (SSA) model for univariate time-series forecasting. For the details of the model, refer to http://arxiv.org/pdf/1206.6910.pdf.
public static Microsoft.ML.Transforms.TimeSeries.SsaForecastingEstimator ForecastBySsa (this Microsoft.ML.ForecastingCatalog catalog, string outputColumnName, string inputColumnName, int windowSize, int seriesLength, int trainSize, int horizon, bool isAdaptive = false, float discountFactor = 1, Microsoft.ML.Transforms.TimeSeries.RankSelectionMethod rankSelectionMethod = Microsoft.ML.Transforms.TimeSeries.RankSelectionMethod.Exact, int? rank = default, int? maxRank = default, bool shouldStabilize = true, bool shouldMaintainInfo = false, Microsoft.ML.Transforms.TimeSeries.GrowthRatio? maxGrowth = default, string confidenceLowerBoundColumn = default, string confidenceUpperBoundColumn = default, float confidenceLevel = 0.95, bool variableHorizon = false);
static member ForecastBySsa : Microsoft.ML.ForecastingCatalog * string * string * int * int * int * int * bool * single * Microsoft.ML.Transforms.TimeSeries.RankSelectionMethod * Nullable<int> * Nullable<int> * bool * bool * Nullable<Microsoft.ML.Transforms.TimeSeries.GrowthRatio> * string * string * single * bool -> Microsoft.ML.Transforms.TimeSeries.SsaForecastingEstimator
<Extension()>
Public Function ForecastBySsa (catalog As ForecastingCatalog, outputColumnName As String, inputColumnName As String, windowSize As Integer, seriesLength As Integer, trainSize As Integer, horizon As Integer, Optional isAdaptive As Boolean = false, Optional discountFactor As Single = 1, Optional rankSelectionMethod As RankSelectionMethod = Microsoft.ML.Transforms.TimeSeries.RankSelectionMethod.Exact, Optional rank As Nullable(Of Integer) = Nothing, Optional maxRank As Nullable(Of Integer) = Nothing, Optional shouldStabilize As Boolean = true, Optional shouldMaintainInfo As Boolean = false, Optional maxGrowth As Nullable(Of GrowthRatio) = Nothing, Optional confidenceLowerBoundColumn As String = Nothing, Optional confidenceUpperBoundColumn As String = Nothing, Optional confidenceLevel As Single = 0.95, Optional variableHorizon As Boolean = false) As SsaForecastingEstimator
Parameters
- catalog
- ForecastingCatalog
Catalog.
- outputColumnName
- String
Name of the column resulting from the transformation of inputColumnName
.
- inputColumnName
- String
Name of column to transform. If set to null
, the value of the outputColumnName
will be used as source.
The vector contains Alert, Raw Score, P-Value as first three values.
- windowSize
- Int32
The length of the window on the series for building the trajectory matrix (parameter L).
- seriesLength
- Int32
The length of series that is kept in buffer for modeling (parameter N).
- trainSize
- Int32
The length of series from the beginning used for training.
- horizon
- Int32
The number of values to forecast.
- isAdaptive
- Boolean
The flag determining whether the model is adaptive.
- discountFactor
- Single
The discount factor in [0,1] used for online updates.
- rankSelectionMethod
- RankSelectionMethod
The rank selection method.
The desired rank of the subspace used for SSA projection (parameter r). This parameter should be in the range in [1, windowSize]. If set to null, the rank is automatically determined based on prediction error minimization.
The maximum rank considered during the rank selection process. If not provided (i.e. set to null), it is set to windowSize - 1.
- shouldStabilize
- Boolean
The flag determining whether the model should be stabilized.
- shouldMaintainInfo
- Boolean
The flag determining whether the meta information for the model needs to be maintained.
- maxGrowth
- Nullable<GrowthRatio>
The maximum growth on the exponential trend.
- confidenceLowerBoundColumn
- String
The name of the confidence interval lower bound column. If not specified then confidence intervals will not be calculated.
- confidenceUpperBoundColumn
- String
The name of the confidence interval upper bound column. If not specified then confidence intervals will not be calculated.
- confidenceLevel
- Single
The confidence level for forecasting.
- variableHorizon
- Boolean
Set this to true if horizon will change after training(at prediction time).
Returns
Examples
using System;
using System.Collections.Generic;
using System.IO;
using Microsoft.ML;
using Microsoft.ML.Transforms.TimeSeries;
namespace Samples.Dynamic
{
public static class Forecasting
{
// This example creates a time series (list of Data with the i-th element
// corresponding to the i-th time slot) and then does forecasting.
public static void Example()
{
// Create a new ML context, for ML.NET operations. It can be used for
// exception tracking and logging, as well as the source of randomness.
var ml = new MLContext();
// Generate sample series data with a recurring pattern.
var data = new List<TimeSeriesData>()
{
new TimeSeriesData(0),
new TimeSeriesData(1),
new TimeSeriesData(2),
new TimeSeriesData(3),
new TimeSeriesData(4),
new TimeSeriesData(0),
new TimeSeriesData(1),
new TimeSeriesData(2),
new TimeSeriesData(3),
new TimeSeriesData(4),
new TimeSeriesData(0),
new TimeSeriesData(1),
new TimeSeriesData(2),
new TimeSeriesData(3),
new TimeSeriesData(4),
};
// Convert data to IDataView.
var dataView = ml.Data.LoadFromEnumerable(data);
// Setup arguments.
var inputColumnName = nameof(TimeSeriesData.Value);
var outputColumnName = nameof(ForecastResult.Forecast);
// Instantiate the forecasting model.
var model = ml.Forecasting.ForecastBySsa(outputColumnName,
inputColumnName, 5, 11, data.Count, 5);
// Train.
var transformer = model.Fit(dataView);
// Forecast next five values.
var forecastEngine = transformer.CreateTimeSeriesEngine<TimeSeriesData,
ForecastResult>(ml);
var forecast = forecastEngine.Predict();
Console.WriteLine($"Forecasted values:");
Console.WriteLine("[{0}]", string.Join(", ", forecast.Forecast));
// Forecasted values:
// [1.977226, 1.020494, 1.760543, 3.437509, 4.266461]
// Update with new observations.
forecastEngine.Predict(new TimeSeriesData(0));
forecastEngine.Predict(new TimeSeriesData(0));
forecastEngine.Predict(new TimeSeriesData(0));
forecastEngine.Predict(new TimeSeriesData(0));
// Checkpoint.
forecastEngine.CheckPoint(ml, "model.zip");
// Load the checkpointed model from disk.
// Load the model.
ITransformer modelCopy;
using (var file = File.OpenRead("model.zip"))
modelCopy = ml.Model.Load(file, out DataViewSchema schema);
// We must create a new prediction engine from the persisted model.
var forecastEngineCopy = modelCopy.CreateTimeSeriesEngine<
TimeSeriesData, ForecastResult>(ml);
// Forecast with the checkpointed model loaded from disk.
forecast = forecastEngineCopy.Predict();
Console.WriteLine("[{0}]", string.Join(", ", forecast.Forecast));
// [1.791331, 1.255525, 0.3060154, -0.200446, 0.5657795]
// Forecast with the original model(that was checkpointed to disk).
forecast = forecastEngine.Predict();
Console.WriteLine("[{0}]", string.Join(", ", forecast.Forecast));
// [1.791331, 1.255525, 0.3060154, -0.200446, 0.5657795]
}
class ForecastResult
{
public float[] Forecast { get; set; }
}
class TimeSeriesData
{
public float Value;
public TimeSeriesData(float value)
{
Value = value;
}
}
}
}
using System;
using System.Collections.Generic;
using System.IO;
using Microsoft.ML;
using Microsoft.ML.Transforms.TimeSeries;
namespace Samples.Dynamic
{
public static class ForecastingWithConfidenceInternal
{
// This example creates a time series (list of Data with the i-th element
// corresponding to the i-th time slot) and then does forecasting.
public static void Example()
{
// Create a new ML context, for ML.NET operations. It can be used for
// exception tracking and logging, as well as the source of randomness.
var ml = new MLContext();
// Generate sample series data with a recurring pattern.
var data = new List<TimeSeriesData>()
{
new TimeSeriesData(0),
new TimeSeriesData(1),
new TimeSeriesData(2),
new TimeSeriesData(3),
new TimeSeriesData(4),
new TimeSeriesData(0),
new TimeSeriesData(1),
new TimeSeriesData(2),
new TimeSeriesData(3),
new TimeSeriesData(4),
new TimeSeriesData(0),
new TimeSeriesData(1),
new TimeSeriesData(2),
new TimeSeriesData(3),
new TimeSeriesData(4),
};
// Convert data to IDataView.
var dataView = ml.Data.LoadFromEnumerable(data);
// Setup arguments.
var inputColumnName = nameof(TimeSeriesData.Value);
var outputColumnName = nameof(ForecastResult.Forecast);
// Instantiate the forecasting model.
var model = ml.Forecasting.ForecastBySsa(outputColumnName,
inputColumnName, 5, 11, data.Count, 5,
confidenceLevel: 0.95f,
confidenceLowerBoundColumn: "ConfidenceLowerBound",
confidenceUpperBoundColumn: "ConfidenceUpperBound");
// Train.
var transformer = model.Fit(dataView);
// Forecast next five values.
var forecastEngine = transformer.CreateTimeSeriesEngine<TimeSeriesData,
ForecastResult>(ml);
var forecast = forecastEngine.Predict();
PrintForecastValuesAndIntervals(forecast.Forecast, forecast
.ConfidenceLowerBound, forecast.ConfidenceUpperBound);
// Forecasted values:
// [1.977226, 1.020494, 1.760543, 3.437509, 4.266461]
// Confidence intervals:
// [0.3451088 - 3.609343] [-0.7967533 - 2.83774] [-0.058467 - 3.579552] [1.61505 - 5.259968] [2.349299 - 6.183623]
// Update with new observations.
forecastEngine.Predict(new TimeSeriesData(0));
forecastEngine.Predict(new TimeSeriesData(0));
forecastEngine.Predict(new TimeSeriesData(0));
forecastEngine.Predict(new TimeSeriesData(0));
// Checkpoint.
forecastEngine.CheckPoint(ml, "model.zip");
// Load the checkpointed model from disk.
// Load the model.
ITransformer modelCopy;
using (var file = File.OpenRead("model.zip"))
modelCopy = ml.Model.Load(file, out DataViewSchema schema);
// We must create a new prediction engine from the persisted model.
var forecastEngineCopy = modelCopy.CreateTimeSeriesEngine<
TimeSeriesData, ForecastResult>(ml);
// Forecast with the checkpointed model loaded from disk.
forecast = forecastEngineCopy.Predict();
PrintForecastValuesAndIntervals(forecast.Forecast, forecast
.ConfidenceLowerBound, forecast.ConfidenceUpperBound);
// [1.791331, 1.255525, 0.3060154, -0.200446, 0.5657795]
// Confidence intervals:
// [0.1592142 - 3.423448] [-0.5617217 - 3.072772] [-1.512994 - 2.125025] [-2.022905 - 1.622013] [-1.351382 - 2.482941]
// Forecast with the original model(that was checkpointed to disk).
forecast = forecastEngine.Predict();
PrintForecastValuesAndIntervals(forecast.Forecast,
forecast.ConfidenceLowerBound, forecast.ConfidenceUpperBound);
// [1.791331, 1.255525, 0.3060154, -0.200446, 0.5657795]
// Confidence intervals:
// [0.1592142 - 3.423448] [-0.5617217 - 3.072772] [-1.512994 - 2.125025] [-2.022905 - 1.622013] [-1.351382 - 2.482941]
}
static void PrintForecastValuesAndIntervals(float[] forecast, float[]
confidenceIntervalLowerBounds, float[] confidenceIntervalUpperBounds)
{
Console.WriteLine($"Forecasted values:");
Console.WriteLine("[{0}]", string.Join(", ", forecast));
Console.WriteLine($"Confidence intervals:");
for (int index = 0; index < forecast.Length; index++)
Console.Write($"[{confidenceIntervalLowerBounds[index]} -" +
$" {confidenceIntervalUpperBounds[index]}] ");
Console.WriteLine();
}
class ForecastResult
{
public float[] Forecast { get; set; }
public float[] ConfidenceLowerBound { get; set; }
public float[] ConfidenceUpperBound { get; set; }
}
class TimeSeriesData
{
public float Value;
public TimeSeriesData(float value)
{
Value = value;
}
}
}
}