ml4k 0.8

# Python wrapper for the machinelearningforkids.co.uk API

This library was designed for use with [machinelearningforkids.co.uk](https://machinelearningforkids.co.uk).

## Installation

Use pip to install the `ml4k` package:

sudo pip install ml4k

## Usage

In order to use your machine learning model, create a `Model` object
with your project's API key:

import ml4k

API_KEY = "PASTE-API-KEY-HERE"
model = ml4k.Model(API_KEY)


The following methods are available:

### `model.classify(data)`

Classifies the given data and returns a dictionary with the result. The
result dictionary contains the `"label"` and `"confidence"` values.


### `model.add_training_data(label, data)`

Adds training data to the given label "bucket". You can pass text,
images, or a list of numbers.


## Examples

### Recognizing Text

# Get input from user and pass it to our model
command = input("Type a command: ")
result = model.classify("Turn on the lamp")

if result["label"] == "lamp_on":
print("Turning on the lamp")
elif result["label"] == "lamp_off":
print("Turning off the lamp")


### Recognizing Images

Recognizing images is similar, but you need to pass binary image data.
There are many was to get image data in python, such as opening and
reading a file, or capturing from the webcam using a the
[SimpleCV](http://simplecv.org/) library.

Images will be automatically downsized before sending to the API.

# Assuming you have image data stored as a binary string...
result = model.classify(image)
if result["label"] == "dog":
print("That's a dog")
elif result["label"] == "cat":
print("That's a cat")