Class TsUtils
java.lang.Object
edu.cmu.tetrad.search.utils.TsUtils
Contains some utilities for doing autoregression. Should probably be improved by somebody.
- Version:
- $Id: $Id
- Author:
- josephramsey, danielmalinsky (some improvements)
-
Nested Class Summary
Nested ClassesModifier and TypeClassDescriptionstatic classGives a result consisting of the residuals and collapsed var graphs. -
Method Summary
Modifier and TypeMethodDescriptionstatic DataSetCreates new time series dataset from the given one with index variable (e.g., time)static booleanallEigenvaluesAreSmallerThanOneInModulus.static DataSetar.static DataSetar2.static DataSetcreateLagData(DataSet data, int numLags) Creates new time series dataset from the given one (fixed to deal with mixed datasets)static DataSetcreateLagData(DataSet data, int numLags, Knowledge knowledge) Creates a new time series dataset from the given one, expanding the provided within-lag Knowledge across all time lags.static DataSetcreateShiftedData(DataSet data, int[] shifts) createShiftedData.static DataSetdifference(DataSet data, int d) Calculates the dth difference of the given data.static intgetIndex.static KnowledgegetKnowledge(Graph graph) getKnowledge.static intgetLag.static StringgetNameNoLag(Object obj) getNameNoLag.static StringgetPrefix.static double[]getSelfLoopCoefs(DataSet timeSeries) getSelfLoopCoefs.static TimeLagGraphgraphToLagGraph(Graph _graph, int numLags, int extraLaggedEdges) Converts a directed graph into a time-lagged graph representation.static TsUtils.VarResultstructuralVar(DataSet timeSeries, int numLags) structuralVar.static doublesumOfArCoefficients(DataSet timeSeries, int numLags) sumOfArCoefficients.
-
Method Details
-
ar
ar.
- Parameters:
timeSeries- aDataSetobjectnumLags- a int- Returns:
- the VAR residuals of the given time series with the given number of lags. That is, every variable at the model lag is regressed onto every variable at previous lags, up to the given number of lags, and the residuals of these regressions for each variable are returned.
-
ar2
-
structuralVar
public static TsUtils.VarResult structuralVar(DataSet timeSeries, int numLags) throws InterruptedException structuralVar.
- Parameters:
timeSeries- aDataSetobjectnumLags- a int- Returns:
- a
TsUtils.VarResultobject - Throws:
InterruptedException- if any
-
createShiftedData
-
getSelfLoopCoefs
-
sumOfArCoefficients
-
difference
Calculates the dth difference of the given data. If d = 0, the original data is returned. If d = 1, the data (with one fewer rows) is returned, with each row subtracted from its successor. If d = 1, the same operation is applied to the result of d = 1. And so on.- Parameters:
data- the data to be differenced.d- the number of differences to take, >= 0.- Returns:
- the differenced data.
-
createLagData
-
createLagData
Creates a new time series dataset from the given one, expanding the provided within-lag Knowledge across all time lags. If the input knowledge has k non-empty tiers, the expanded knowledge will have k * (numLags + 1) tiers, with the within-lag tier structure (including forbidden/required edges and tierForbiddenWithin flags) replicated at each lag. Cross-lag temporal ordering (no edges from later time back to earlier time) is enforced automatically by the tier structure.- Parameters:
data- the time series datasetnumLags- the number of lagsknowledge- within-lag knowledge over base variable names (no lag suffix)- Returns:
- the lagged dataset with expanded knowledge attached
-
addIndex
-
graphToLagGraph
Converts a directed graph into a time-lagged graph representation. The resulting graph includes lagged copies of nodes, self-lagged edges, contemporaneous edges, and probabilistically generated inter-variable lagged edges up to the specified maximum lag.- Parameters:
_graph- The base DAG graph to be converted. Must only contain directed edges. Throws IllegalArgumentException if the input graph is not fully directed.numLags- The maximum number of lags to include in the time-lagged graph.extraLaggedEdges- The number of extra lagged edges to include in the time-lagged graph.- Returns:
- A time-lagged graph representation of the input graph with the specified properties. The resulting graph includes lagged nodes, self-lagged edges, and probabilistically generated lagged edges.
-
getNameNoLag
-
getPrefix
-
getIndex
-
getLag
-
getKnowledge
-
allEigenvaluesAreSmallerThanOneInModulus
-