Confusion Table Transpose Convention Tool

Description

Retain it when rowsAreActual or transpose it otherwise, returning actualRowsPredictedColumns.

A classification summary depends on which axis represents actual labels, which represents predictions, and how a decision threshold handles equality. A local matrix transformation or threshold check does not measure overall model quality.

Retain it when rowsAreActual or transpose it otherwise, returning actualRowsPredictedColumns.

When to use this tool

Use Confusion Table Transpose Convention when you need one explicit classification-record component.

How to use the tool and how it works

Enter counts, rowsAreActual. Use the units and conventions below.
Run the tool. Correct any format, range, or cross-field input error before interpreting the output.
Read actualRowsPredictedColumns. Use the output convention below.

Require a nonempty square matrix of nonnegative integer counts. Retain it when rowsAreActual or transpose it otherwise, returning actualRowsPredictedColumns.^[1]

Input field	Meaning and units	Accepted input
counts	Counts	Required. Array of array of finite number; maximum items 128; maximum items 128.
rowsAreActual	Rows Are Actual	Required. Boolean.

Output field	Interpretation	Returned shape
actualRowsPredictedColumns	Actual Rows Predicted Columns	Required. Array of array of finite number; maximum items 128; maximum items 128.

json

{
  "input": {
    "counts": [
      [
        12,
        3
      ],
      [
        2,
        8
      ]
    ],
    "rowsAreActual": false
  },
  "output": {
    "actualRowsPredictedColumns": [
      [
        12,
        2
      ],
      [
        3,
        8
      ]
    ]
  }
}

Limitations and assumptions

Class order and labels must already agree; no predictions, class identity, normalization, or accuracy metrics are evaluated.
Inputs must satisfy both the field types and the cross-field conditions above. Invalid input causes an error; a successful false flag or documented placeholder is a different result.

Alternative or Complementary methods

Evaluate the complete prediction set and report suitable metrics under a documented label and threshold convention.

References

Calculation and input contract: classification/confusion-table-transpose-convention — tool.io

Similar or alternative tools

Matrix Transpose Shape Pair
Require nonnegative integer rows and columns and swap them for transposedRows and transposedColumns.
Classifier Threshold Equality Policy
Return positive for score above threshold or exact equality when equalityIsPositive.