Python Library

Python library for Akkio

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.

Installation

pip install akkio

Example Usage

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

# list models in your organization
models = akkio.get_models()['models']
for model in models:
  print(model)

# list datasets in your organization
datasets = akkio.get_datasets()['datasets']
for dataset in datasets:
  print(dataset)

# create a new empty dataset
new_dataset = akkio.create_dataset('python api test')
print(new_dataset)

# add rows to the 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)

# create a model
new_model = akkio.create_model(new_dataset['dataset_id'], ['y'], [], {'duration': 1})
print(new_model)

# make a prediction using the model
prediction = akkio.make_prediction(new_model['model_id'], [{'x': 0.1}, {'x':0.7}], explain=True)
print(prediction)

Datasets

create_dataset(dataset_name)

Create a new empty dataset.

input	description
dataset_name	The name of your newly created dataset.

add_rows_to_dataset(dataset_id, rows)

Add rows to a dataset.

input	description
dataset_id	A dataset id
rows	An array of rows to be added to the dataset in the following form: [{ "field 1": "data", "field 2": "data" }, { ... }, ... ]

get_datasets()

Get all datasets in your organization.

get_dataset(dataset_id)

Get a dataset.

input	description
dataset_id	A dataset id

parse_dataset(dataset_id)

Recalculate the field types for a dataset.

input	description
dataset_id	A dataset id

delete_dataset(dataset_id)

Delete a dataset.

input	description
dataset_id	A dataset id

Models

get_models()

Get all models in your organization.

delete_model(model_id)

Delete a model in your organization.

input	description
model_id	A model id

create_model(dataset_id, predict_fields, ignore_fields, params)

Create a model (requires a dataset).

input	description
dataset_id	A dataset id
predict_fields	An array of field names to predict (case sensitive)
ignore_fields	An array of field names to ignore (case sensitive) (optional)
params	A dict with default value of: { "duration": 10, "extra_attention: False, "force": False } duration is the duration in seconds to be used for model training. extra_attention can be enabled to help with predicting rare cases force forces a new model to be created

Sometimes creating models can take a while, especially if this is the first time creating a model on this dataset. create_model is idempotent and can be called multiple times with the same parameters.

make_prediction(model_id, data)

Make a prediction using your model and new data.

input	description
model_id	A model id
data	An array of rows to be predicted in the following form: [{ "field 1": "data", "field 2": "data" }, { ... }, ... ]