Namespace: statistics

This namespace holds various functions useful for basic statistics and analysis of trees and forests

Functions

getMedian

getMedian(numbers): number

Parameters

Name	Type
`numbers`	`number`[]

Returns

number

Defined in

statistic/basicStatistic.ts:3

getArithmeticAverage

getArithmeticAverage(values): number

Parameters

Name	Type
`values`	`number`[]

Returns

number

Defined in

statistic/basicStatistic.ts:19

getMostCommonValue

getMostCommonValue(values): string

Parameters

Name	Type
`values`	`string`[]

Returns

string

Defined in

statistic/getMostCommonValue.ts:28

getMostCommonValues

getMostCommonValues(values): string[]

returns most common values of input array (if counts are equal return all values with this count)

Parameters

Name	Type
`values`	`string`[]

Returns

string[]

Defined in

statistic/getMostCommonValue.ts:6

getStandardDeviation

getStandardDeviation(values): number

Parameters

Name	Type
`values`	`number`[]

Returns

number

Defined in

statistic/basicStatistic.ts:35

getVariance

getVariance(values): number

Parameters

Name	Type
`values`	`number`[]

Returns

number

Defined in

statistic/basicStatistic.ts:27

getMissClassificationRate

getMissClassificationRate(treeRootNode, dataSet, configuration): number

Wrapper for getMissClassificationRateRaw

Parameters

Name	Type
`treeRootNode`	`TreeGardenNode`
`dataSet`	`TreeGardenDataSample`[]
`configuration`	`TreeGardenConfiguration`

Returns

number

Defined in

statistic/treeStats.ts:72

getMissClassificationRateRaw

getMissClassificationRateRaw(realClasses, predictedClasses): number

Used for calculation of accuracy of classification trees, number between 0 and 1

0 - 0% of correct classifications
1 - 100% of correct classifications

Parameters

Name	Type
`realClasses`	(`undefined` \| `string` \| `number`)[]
`predictedClasses`	(`undefined` \| `string` \| `number`)[]

Returns

number

Defined in

statistic/treeStats.ts:59

getRAbsError

getRAbsError(treeRootNode, dataSet, configuration): number

Wrapper for getRAbsErrorRaw

Parameters

Name	Type
`treeRootNode`	`TreeGardenNode`
`dataSet`	`TreeGardenDataSample`[]
`configuration`	`TreeGardenConfiguration`

Returns

number

Defined in

statistic/treeStats.ts:40

getRAbsErrorRaw

getRAbsErrorRaw(realValues, predictedValues): number

Used for calculation of accuracy of regression trees, number up to 1

tree-garden implements modified coefficient of determination, it uses absolute values instead of squared values.

Remarks

Do not be scared by negative values - as number 0 means - model predicts as good as average value of your data set - comparison to base model. If model predicts worse, than average value, it will be negative. Ideal model will have 1.

Parameters

Name	Type
`realValues`	`number`[]
`predictedValues`	`number`[]

Returns

number

Defined in

statistic/treeStats.ts:20

getNumberOfSamplesInNode

getNumberOfSamplesInNode(node): number

Parameters

Name	Type
`node`	`TreeGardenNode`

Returns

number

Defined in

statistic/treeStats.ts:108

getNumberOfTreeNodes

getNumberOfTreeNodes(treeRoot): number

Parameters

Name	Type
`treeRoot`	`TreeGardenNode`

Returns

number

Defined in

statistic/treeStats.ts:107

getTreeDepth

getTreeDepth(tree): number

Parameters

Name	Type
`tree`	`TreeGardenNode`

Returns

number

Defined in

statistic/treeStats.ts:111

getOutOfTheBagError

getOutOfTheBagError(treesAndOutOfTheBagSets, fullDataSet, config, majorityVotingFn?): number

Function for calculation of out of the bag error for random forest. It is calculated by default, during training of random forest. See random forest example

Parameters

Name	Type	Default value
`treesAndOutOfTheBagSets`	[`TreeGardenNode`, `Set`<`undefined` \| `string` \| `number`>][]	`undefined`
`fullDataSet`	`TreeGardenDataSample`[]	`undefined`
`config`	`TreeGardenConfiguration`	`undefined`
`majorityVotingFn`	(`treeRoots`: `TreeGardenNode`[], `dataSample`: `TreeGardenDataSample`, `config`: `TreeGardenConfiguration`) => `SingleSamplePredictionResult`	`getResultFromMultipleTrees`

Returns

number

Defined in

statistic/randomForestStats.ts:14