stanfordnlp/CoreNLP · error · edu.stanford.nlp.io.RuntimeIOException

Caught IOException outputting QN data to file

Error message

Caught IOException outputting QN data to file

What it means

When QNMinimizer is configured to write output to files (outFile/infoFile), any IOException while opening or writing these files is wrapped in RuntimeIOException so it escapes as an unchecked exception. It signals failure to create the diagnostic/output files, not a problem with the optimization itself.

Solutions

1. Ensure the output directory exists and is writable before calling minimize().
2. Catch RuntimeIOException around minimize() and handle/log it.
3. Disable or relocate file output if it is not needed.

Example fix

// before
File dir = new File("/logs/qn");
minimizer.minimize(f, tol, init); // RuntimeIOException if /logs/qn missing
// after
File dir = new File("/logs/qn");
dir.mkdirs();
minimizer.setOutputFile(new File(dir, "qn-out"));
minimizer.minimize(f, tol, init);

File outDir = new File(baseName).getAbsoluteFile().getParentFile();
if (outDir == null || !outDir.isDirectory() || !outDir.canWrite())
  throw new IllegalStateException("QN output directory not writable: " + outDir);

try {
  minimizer.minimize(f, tol, init);
} catch (RuntimeIOException e) {
  log.severe("QN output IO failed: " + e.getMessage());
  // retry without file output or with a new directory
}

When it happens

Trigger: Calling minimize() with output writing enabled (e.g. QNMinimizer set to dump function/gradient info) when the base output path is unwritable, the directory does not exist, disk is full, or a FileOutputStream to baseName + ".txt"/".info" fails.

Common situations: Running training on a read-only filesystem, a nonexistent output directory, insufficient write permissions, or reusing a path that is locked by another process.

Understand the failure class

Background: "failed to write file", "Could not save figure", "Error saving remote file" — file write failed: causes and fixes across languages and libraries — this error's family across 38 libraries.

Related errors

• Cannot initialize logger!
• Could not create output stream (cannot write file): " +…
• Error creating data exporter
• Error opening output file
• Exception: in printToFile

AI-assisted analysis of stanfordnlp/CoreNLP@1b7edd19c4 (2026-09-10). Data as JSON: /api/errors/22ecc6449103b355. Report an issue: GitHub.

Appendix: source

Thrown at src/edu/stanford/nlp/optimization/QNMinimizer.java:934

    } else {
      grad = rawGrad;
    }

    PrintWriter outFile = null;
    PrintWriter infoFile = null;

    if (outputToFile) {
      try {
        String baseName = "QN_m" + mem + '_' + lsOpt.toString() + '_'
            + scaleOpt.toString();
        outFile = new PrintWriter(new FileOutputStream(baseName + ".output"),
            true);
        infoFile = new PrintWriter(new FileOutputStream(baseName + ".info"),
            true);
        infoFile.println(dFunction.domainDimension() + "; DomainDimension ");
        infoFile.println(mem + "; memory");
      } catch (IOException e) {
        throw new RuntimeIOException("Caught IOException outputting QN data to file", e);
      }
    }

    Record rec = new Record(monitor, functionTolerance, outFile);
    // sets the original gradient and x. Also stores the monitor.
    rec.start(value, rawGrad, x);

    // Check if max Evaluations and Iterations have been provided.
    maxFevals = (maxFunctionEvaluations > 0) ? maxFunctionEvaluations
        : Integer.MAX_VALUE;
    // maxIterations = (maxIterations > 0) ? maxIterations : Integer.MAX_VALUE;

    if (!quiet) {
      log.info("               An explanation of the output:");
      log.info("Iter           The number of iterations");
      log.info("evals          The number of function evaluations");
      log.info("SCALING        <D> Diagonal scaling was used; <I> Scaled Identity");
      log.info("LINESEARCH     [## M steplength]  Minpack linesearch");

View on GitHub (pinned to 1b7edd19c4)