The ML.CONFUSION_MATRIX function

ML.CONFUSION_MATRIX function

Use the ML.CONFUSION_MATRIX function to return a confusion matrix for the input classification model and input data.

For information about model evaluation in BigQuery ML, see Model evaluation overview.

For information about supported model types of each SQL statement and function, and all supported SQL statements and functions for each model type, read End-to-end user journey for each model.

ML.CONFUSION_MATRIX syntax

ML.CONFUSION_MATRIX(MODEL model_name
           [, {TABLE table_name | (query_statement)}]
           [, STRUCT(<T> AS threshold)])

model_name

model_name is the name of the model you're evaluating. If you do not have a default project configured, prepend the project ID to the model name in following format: `[PROJECT_ID].[DATASET].[MODEL]` (including the backticks); for example, `myproject.mydataset.mymodel`.

table_name

(Optional) table_name is the name of the input table that contains the evaluation data. If you do not have a default project configured, prepend the project ID to the table name in following format: `[PROJECT_ID].[DATASET].[TABLE]` (including the backticks); for example, `myproject.mydataset.mytable`.

If table_name is specified, the input column names in the table must match the column names in the model, and their types should be compatible according to BigQuery implicit coercion rules. The input must have a column that matches the label column name provided during training. This value is provided using the input_label_cols option. If input_label_cols is unspecified, the column named "label" in the training data is used.

If neither table_name nor query_statement is specified, ML.CONFUSION_MATRIX computes the confusion matrix results as follows:

• If the data is split during training, the split evaluation data is used to compute the confusion matrix results.
• If the data is not split during training, the entire training input is used to compute the confusion matrix results.

Note: ML.CONFUSION_MATRIX requires input data with some older models, and will return an error if it is absent. If so, provide input data when using ML.CONFUSION_MATRIX with these models.

query_statement

(Optional) The query_statement clause specifies the GoogleSQL query that is used to generate the evaluation data. See the GoogleSQL Query Syntax page for the supported SQL syntax of the query_statement clause.

If query_statement is specified, the input column names from the query must match the column names in the model, and their types should be compatible according to BigQuery implicit coercion rules. The input must have a column that matches the label column name provided during training. This value is provided using the input_label_cols option. If input_label_cols is unspecified, the column named "label" in the training data is used. The extra columns are ignored.

If the TRANSFORM clause was present in the CREATE MODEL statement that created model_name, then only the input columns present in the TRANSFORM clause must appear in query_statement.

If neither table_name nor query_statement is specified, ML.CONFUSION_MATRIX computes the confusion matrix results as follows:

• If the data is split during training, the split evaluation data is used to compute the confusion matrix results.
• If the data is not split during training, the entire training input is used to compute the confusion matrix results.

Note: ML.CONFUSION_MATRIX requires input data with some older models, and will return an error if it is absent. If so, provide input data when using ML.CONFUSION_MATRIX with these models.

threshold

(Optional) threshold is the custom threshold for your binary classification model used for evaluation. The default value is 0.5. The threshold value that is supplied must be of type STRUCT.

A zero value for precision or recall means that the selected threshold produced no true positive labels. A NaN value for precision means that the selected threshold produced no positive labels, neither true positives nor false positives.

Note: threshold cannot be used with multiclass classification models.

ML.CONFUSION_MATRIX output

The output columns of the ML.CONFUSION_MATRIX function depend on the model. The first output column is always expected_label. There are N additional columns, one for each class in the trained model. The names of the additional columns depend on the class labels used to train the model.

If the training class labels all conform to BigQuery column naming rules, the labels are used as the column names. Columns that do not conform to naming rules are altered to conform to the column naming rules and to be unique. For example, if the labels are 0 and 1, the output column names are _0 and _1.

The columns are ordered based on the class labels in ascending order. If the labels in the evaluation data match those in the training data, the True Positives are shown on the diagonal from top left to bottom right. The expected (or actual) labels are listed one per row, and the predicted labels are listed one per column.

The values in the expected_label column are the exact values and type passed into ML.CONFUSION_MATRIX in the label column of the evaluation data. This is true even if they do not exactly match the values or type used during training.

ML.CONFUSION_MATRIX example

The following example assumes your model and input table are in your default project.

The following query is used to return the confusion matrix for a logistic regression model.

SELECT
  *
FROM
  ML.CONFUSION_MATRIX(MODEL `mydataset.mymodel`,
  (
    SELECT
      *
    FROM
      `mydataset.mytable`))

ML.CONFUSION_MATRIX limitations

The ML.CONFUSION_MATRIX function is subject to the following limitations:

• ML.CONFUSION_MATRIX does not support imported TensorFlow models.