Liking cljdoc? Tell your friends :D

scicloj.metamorph.ml.verify


basic-classificationclj

(basic-classification options-map)
(basic-classification options-map max-avg-loss)

Verifies a classification model performs within acceptable error bounds.

options-map - Model options map (must include :model-type) max-avg-loss - Maximum acceptable average MAE (default: 0.5)

Trains the model 5 times on Titanic survival data, calculates average MAE on test set, and asserts it's below max-avg-loss.

Returns a clojure.test assertion result.

Used for testing model implementations and ensuring classification models work correctly.

Verifies a classification model performs within acceptable error bounds.

`options-map` - Model options map (must include `:model-type`)
`max-avg-loss` - Maximum acceptable average MAE (default: 0.5)

Trains the model 5 times on Titanic survival data, calculates average MAE on
test set, and asserts it's below `max-avg-loss`.

Returns a clojure.test assertion result.

Used for testing model implementations and ensuring classification models work correctly.
sourceraw docstring

basic-regressionclj

(basic-regression options-map)
(basic-regression options-map max-avg-loss)

Verifies a regression model performs within acceptable error bounds.

options-map - Model options map (must include :model-type) max-avg-loss - Maximum acceptable average MAE (default: 0.5)

Trains the model 5 times on Iris data (predicting petal width), calculates average MAE on test set, and asserts it's below max-avg-loss.

Returns a clojure.test assertion result.

Used for testing model implementations and ensuring regression models work correctly.

Verifies a regression model performs within acceptable error bounds.

`options-map` - Model options map (must include `:model-type`)
`max-avg-loss` - Maximum acceptable average MAE (default: 0.5)

Trains the model 5 times on Iris data (predicting petal width), calculates
average MAE on test set, and asserts it's below `max-avg-loss`.

Returns a clojure.test assertion result.

Used for testing model implementations and ensuring regression models work correctly.
sourceraw docstring

classification-titanic*clj

Lazy-loaded Titanic dataset configured for classification testing.

A delayed dataset with missing values dropped, categorical columns converted to numeric, and :survived set as the inference target. Used by basic-classification for model verification.

Deref with @classification-titanic* to load.

Lazy-loaded Titanic dataset configured for classification testing.

A delayed dataset with missing values dropped, categorical columns converted
to numeric, and `:survived` set as the inference target. Used by
`basic-classification` for model verification.

Deref with `@classification-titanic*` to load.
sourceraw docstring

regression-iris*clj

Lazy-loaded Iris dataset configured for regression testing.

A delayed dataset with species removed and :petal-width set as the inference target. Used by basic-regression for model verification.

Deref with @regression-iris* to load.

Lazy-loaded Iris dataset configured for regression testing.

A delayed dataset with species removed and `:petal-width` set as the inference
target. Used by `basic-regression` for model verification.

Deref with `@regression-iris*` to load.
sourceraw docstring

cljdoc builds & hosts documentation for Clojure/Script libraries

Keyboard shortcuts
Ctrl+kJump to recent docs
Move to previous article
Move to next article
Ctrl+/Jump to the search field
× close