stanfordnlp/CoreNLP · error · RuntimeException

Dataset could not be loaded

Error message

Dataset could not be loaded

What it means

readDataset wraps its entire file-reading loop in a try/catch and rethrows any Exception as a RuntimeException("Dataset could not be loaded", e) with the original as the cause. This generic wrapper can mask the real problem — an IO error, NumberFormatException during feature parsing, or the line-format errors above — so always inspect getCause().

Solutions

  1. Read the cause via e.getCause() / print the full stack trace to find the underlying failure
  2. Verify the dataset file path exists and is readable
  3. Check that all feature values in the file parse correctly for the configured column types

Example fix

// before
Pair<Dataset<String,String>, List<String[]>> p = cdc.readDataset(filename);
// after
try { var p = cdc.readDataset(filename); }
catch (RuntimeException e) { e.getCause().printStackTrace(); throw e; }
Defensive patterns

Strategy: try-catch

Validate before calling

if (!Files.isReadable(Paths.get(filename))) throw new IllegalStateException("Cannot read dataset file: " + filename);

Type guard

null

Try / catch

try { readDataset(f); } catch (RuntimeException e) { Throwable cause = e.getCause(); if (cause != null) cause.printStackTrace(); }

Prevention

When it happens

Trigger: Any exception while reading the dataset file in readDataset: unreadable file path, IOException from the reader, or parsing failures inside makeDatumFromStrings, all rethrown uniformly.

Common situations: Wrong filename/path passed to -testFile or training data options; unreadable file permissions; feature values that fail to parse into numbers; nested line-format errors from entries 4/5.

Understand the failure class

Background: "failed to read file", EACCES, ENOENT and "could not read <path>" errors: when a program can't read a file from disk — this error's family across 49 libraries.

Related errors

AI-assisted analysis of stanfordnlp/CoreNLP@1b7edd19c4 (2026-09-10). Data as JSON: /api/errors/9557b26b9e63e7a6. Report an issue: GitHub.

Appendix: source

Thrown at src/edu/stanford/nlp/classify/ColumnDataClassifier.java:477

          }
          if (strings.length > maxColumns) {
            maxColumns = strings.length;
          }
          if (inTestPhase) {
            lineInfos.add(strings);
          }
          if (strings.length < flags.length) {
            throw new RuntimeException("Error: Line has too few tab-separated columns (" + maxColumns +
                    ") for " + flags.length + " columns required by specified properties: " + line);
          }
          dataset.add(makeDatumFromStrings(strings));
        }
        if (lineNo > 0 && minColumns != maxColumns) {
          logger.info("WARNING: Number of tab-separated columns in " +
                  filename + " varies between " + minColumns + " and " + maxColumns);
        }
      } catch (Exception e) {
        throw new RuntimeException("Dataset could not be loaded", e);
      }
    }
    logger.info("Reading dataset from " + filename + " ... done [" + tim.toSecondsString() + "s, " + dataset.size() + " items].");
    return new Pair<>(dataset, lineInfos);
  }

  /** Split according to whether we are using tsv file (default) or csv files. */
  private String[] splitLineToFields(String line) {
    if (globalFlags.csvInput) {
      String[] strings = StringUtils.splitOnCharWithQuoting(line, ',', '"', '"');
      for (int i = 0; i < strings.length; ++i) {
        if (strings[i].startsWith("\"") && strings[i].endsWith("\""))
          strings[i] = strings[i].substring(1,strings[i].length()-1);
      }
      return strings;
    }
    else {
      return tab.split(line);

View on GitHub (pinned to 1b7edd19c4)