Quickstart

Get up and running with the Akkio API using one of our convenient libraries.

API Keys

As noted in the code samples below, you must get your API keys and copy them into your API code. Those can be found under the team settings page at the bottom of the Akkio app.

Test Code

Node.js

Installation

npm install akkio --save

Usage

const akkio = require('akkio')('your API key');
#get your API key at https://app.akkio.com/team-settings


(async () => {
  #create a new dataset
  let newDataset = await akkio.createDataset('my new dataset');

  #populate it with some toy data
  let rows = [];
  for (var i = 0; i < 1000; i++) {
okay     let x = Math.random();
    rows.push({
      'x': x,
      'value larger than 0.5': x > 0.5,
    });
  }
  await akkio.addRowsToDataset(newDataset.dataset_id, rows);

  # train a model
  let model = await akkio.createModel(newDataset.dataset_id, ['value larger than 0.5'], [], {
    duration: 1
  });

  ## field importance
  for (let field in model.field_importance) {
    console.log('field:', field, 'importance:', model.field_importance[field]);
  }

  # model stats
  for (let field of model.stats) {
    for (let outcome of field) {
      console.log(outcome);
    }
  }

  # use the trained model to make predictions
  let predictions = await akkio.makePrediction(model.model_id, [{
    'x': 0.25
  }, {
    'x': 0.75
  }], {
    explain: true
  });
  console.log(predictions);

})();

Python

Installation

The akkio library is available on Python 3 and can be installed easily on your python instance. Run the following command or search your available packages for akkio (image also below.

pip install akkio

Usage

import akkio
akkio.api_key = 'YOUR-API-KEY-HERE' 
# get your API key at https://app.akkio.com/team-settings

models = akkio.get_models()['models']
for model in models:
  print(model)

datasets = akkio.get_datasets()['datasets']
for dataset in datasets:
  print(dataset)

new_dataset = akkio.create_dataset('python api test')
print(new_dataset)

# create a toy dataset
import random
rows = []
for i in range(1000):
  rows.append({
    'x': random.random()
  })
  rows[-1]['y'] = rows[-1]['x'] > 0.5

akkio.add_rows_to_dataset(new_dataset['dataset_id'], rows)

new_model = akkio.create_model(new_dataset['dataset_id'], ['y'], [], {'duration': 1})
print(new_model)
prediction = akkio.make_prediction(new_model['model_id'], [{'x': 0.1}, {'x':0.7}], explain=True)
print(prediction)

Once you have installed the Akkio custom package and run the above Usage code you should be able to go to app.akkio.com and see an api test project as shown here.