stanfordnlp/CoreNLP · error · IllegalArgumentException

conditionalLogProbGivenNext requires given one less than…

Error message

conditionalLogProbGivenNext requires given one less than clique size ( ) but was

What it means

conditionalLogProbGivenNext computes the log probability of a label at the END of the window given the NEXT windowSize-1 labels. It validates that `given` has exactly windowSize-1 entries so the given context plus the predicted label fills the clique. Other lengths cannot be mapped onto the table's indices and throw IllegalArgumentException.

Solutions

Pass exactly windowSize-1 following labels, taking windowSize from the FactorTable.
When scanning backwards, slice labels from i+1 to i+windowSize (exclusive), yielding windowSize-1 entries.
Validate lengths defensively before calling FactorTable query methods.

Example fix

// before
int[] given = Arrays.copyOfRange(labels, pos+1, labels.length);
ft.conditionalLogProbGivenNext(given, labels[pos]);

// after
int[] given = Arrays.copyOfRange(labels, pos+1, pos+1 + (ft.windowSize - 1));
ft.conditionalLogProbGivenNext(given, labels[pos]);

Defensive patterns

Strategy: validation

Validate before calling

// Java
int end = Math.min(pos + 1 + (ft.windowSize - 1), labels.length);
int[] given = Arrays.copyOfRange(labels, pos + 1, end);
if (given.length != ft.windowSize - 1)
  throw new IllegalArgumentException("insufficient following context");
double lp = ft.conditionalLogProbGivenNext(given, label);

Try / catch

// Java
try {
  lp = ft.conditionalLogProbGivenNext(given, of);
} catch (IllegalArgumentException e) {
  // rebuild given with exactly windowSize-1 following labels
}

Prevention

For backward scans, slice exactly windowSize-1 labels after the current position.
Guard sequence boundaries — near the end of the sequence there may not be enough following context.
Reuse a shared context-building helper across all FactorTable query methods.

When it happens

Trigger: Calling conditionalLogProbGivenNext(int[] given, int of) with given.length != windowSize-1 — e.g. passing the whole following sequence instead of the next windowSize-1 labels, or reusing an array sized for another order.

Common situations: Bidirectional/backward scanning code over a label sequence that slices the wrong window length; demos/tests (this method is only reached via main per the call graph) with hardcoded label arrays; mixing order-1 and order-2 assumptions.

Understand the failure class

Background: "Must be a positive integer", "Invalid value", "Unsupported": the invalid-argument-value error family, when a library rejects the value you pass — this error's family across 35 libraries.

Related errors

AI-assisted analysis of stanfordnlp/CoreNLP@1b7edd19c4 (2026-09-10). Data as JSON: /api/errors/6a04c6805e0d6187. Report an issue: GitHub.

Appendix: source

Thrown at src/edu/stanford/nlp/ie/crf/FactorTable.java:360

    // compute P(given)
    // double probGiven = logProbFront(given);
    // double probGiven = unnormalizedLogProbFront(given);

    // compute P(given, of) / P(given)
    // return probAll - probGiven;
    return probAll;
  }

  /**
   * Computes the probability of the tag OF being at the beginning of the table
   * given that the tag sequence GIVEN is at the end of the table. given is at
   * the end, of is at the beginning
   *
   * @return the probability of the tag of being at the beginning of the table
   */
  public double conditionalLogProbGivenNext(int[] given, int of) {
    if (given.length != windowSize - 1) {
      throw new IllegalArgumentException("conditionalLogProbGivenNext requires given one less than clique size (" +
          windowSize + ") but was " + Arrays.toString(given));
    }
    int[] label = indicesEnd(given);
    double[] masses = new double[label.length];
    for (int i = 0; i < masses.length; i++) {
      masses[i] = table[label[i]];
    }
    return table[indexOf(of, given)] - ArrayMath.logSum(masses);
  }

  public double unnormalizedLogProbFront(int[] labels) {
    int startIndex = indicesFront(labels);
    int numCellsToSum = SloppyMath.intPow(numClasses, windowSize - labels.length);
    // double[] masses = new double[labels.length];
    // for (int i = 0; i < masses.length; i++) {
    //   masses[i] = table[labels[i]];
    // }
    return ArrayMath.logSum(table, startIndex, startIndex + numCellsToSum);

View on GitHub (pinned to 1b7edd19c4)