create docs for StratifiedRandomSplit

This commit is contained in:
Arkadiusz Kondas 2016-07-11 00:07:07 +02:00
parent f04cc04da5
commit ee6ea3b850
4 changed files with 48 additions and 2 deletions

View File

@ -50,6 +50,7 @@ composer require php-ai/php-ml
* [Accuracy](http://php-ml.readthedocs.io/en/latest/machine-learning/metric/accuracy/)
* Cross Validation
* [Random Split](http://php-ml.readthedocs.io/en/latest/machine-learning/cross-validation/random-split/)
* [Stratified Random Split](http://php-ml.readthedocs.io/en/latest/machine-learning/cross-validation/stratified-random-split/)
* Preprocessing
* [Normalization](http://php-ml.readthedocs.io/en/latest/machine-learning/preprocessing/normalization/)
* [Imputation missing values](http://php-ml.readthedocs.io/en/latest/machine-learning/preprocessing/imputation-missing-values/)

View File

@ -50,6 +50,7 @@ composer require php-ai/php-ml
* [Accuracy](http://php-ml.readthedocs.io/en/latest/machine-learning/metric/accuracy/)
* Cross Validation
* [Random Split](http://php-ml.readthedocs.io/en/latest/machine-learning/cross-validation/random-split/)
* [Stratified Random Split](http://php-ml.readthedocs.io/en/latest/machine-learning/cross-validation/stratified-random-split/)
* Preprocessing
* [Normalization](http://php-ml.readthedocs.io/en/latest/machine-learning/preprocessing/normalization/)
* [Imputation missing values](http://php-ml.readthedocs.io/en/latest/machine-learning/preprocessing/imputation-missing-values/)

View File

@ -6,7 +6,7 @@ One of the simplest methods from Cross-validation is implemented as `RandomSpilt
* $dataset - object that implements `Dataset` interface
* $testSize - a fraction of test split (float, from 0 to 1, default: 0.3)
* $seed - seed for random generator (for tests)
* $seed - seed for random generator (e.g. for tests)
```
$randomSplit = new RandomSplit($dataset, 0.2);

View File

@ -0,0 +1,44 @@
# Stratified Random Split
Analogously to `RandomSpilt` class samples are split to two groups: train group and test group.
Distribution of samples takes into account their targets and trying to divide them equally.
You can adjust number of samples in each group.
### Constructor Parameters
* $dataset - object that implements `Dataset` interface
* $testSize - a fraction of test split (float, from 0 to 1, default: 0.3)
* $seed - seed for random generator (e.g. for tests)
```
$split = new StratifiedRandomSplit($dataset, 0.2);
```
### Samples and labels groups
To get samples or labels from test and train group you can use getters:
```
$dataset = new StratifiedRandomSplit($dataset, 0.3, 1234);
// train group
$dataset->getTrainSamples();
$dataset->getTrainLabels();
// test group
$dataset->getTestSamples();
$dataset->getTestLabels();
```
### Example
```
$dataset = new ArrayDataset(
$samples = [[1], [2], [3], [4], [5], [6], [7], [8]],
$targets = ['a', 'a', 'a', 'a', 'b', 'b', 'b', 'b']
);
$split = new StratifiedRandomSplit($dataset, 0.5);
```
Split will have equals amount of each target. Two of the target `a` and two of `b`.