# gizai:ml/predict

prediction with a trained model.

```gizlang
import * as Predict from "gizai:ml/predict";
```

## Functions

### `partOutputs`

```gizlang
function partOutputs(e: Estimator, row: Vec<f64>, d: i64, out: mut Vec<f64>): void
```

Appends the raw outputs of estimator `e` for one row of `d` features to `out`.

Effects: fail.

### `outputsRows`

```gizlang
function outputsRows(m: Model, x: Vec<f64>, from: i64, to: i64): Vec<Vec<f64>>
```

The raw outputs of rows from..to of x, one Vec for each row.

Effects: fail.

### `classIndex`

```gizlang
function classIndex(m: Model, s: Vec<f64>): i64
```

The class index of raw outputs.

Effects: fail.

### `probabilities`

```gizlang
function probabilities(m: Model, s: Vec<f64>): Vec<f64>
```

The class probabilities of model `m` for raw scores `s`, by its link.

Effects: fail.

### `rowParts`

```gizlang
function rowParts(n: i64): i64
```

How many parts `n` rows are predicted in (Task.parallel spreads them on the threads): one part each 1,024 rows, 64 at most.

Effects: fail.

gizai:ml/predict

prediction with a trained model.

gizlang

import * as Predict from "gizai:ml/predict";

Functions

`partOutputs`

gizlang

function partOutputs(e: Estimator, row: Vec<f64>, d: i64, out: mut Vec<f64>): void

Appends the raw outputs of estimator e for one row of d features to out.

Effects: fail.

`outputsRows`

gizlang

function outputsRows(m: Model, x: Vec<f64>, from: i64, to: i64): Vec<Vec<f64>>

The raw outputs of rows from..to of x, one Vec for each row.

Effects: fail.

`classIndex`

gizlang

function classIndex(m: Model, s: Vec<f64>): i64

The class index of raw outputs.

Effects: fail.

`probabilities`

gizlang

function probabilities(m: Model, s: Vec<f64>): Vec<f64>

The class probabilities of model m for raw scores s, by its link.

Effects: fail.

`rowParts`

gizlang

function rowParts(n: i64): i64

How many parts n rows are predicted in (Task.parallel spreads them on the threads): one part each 1,024 rows, 64 at most.

Effects: fail.