AIoT Studio

AIoT Studio is the official library to develop datasources, notebooks and dependency sets for mnubo's AIoT Studio.

Installation

Requirements

• Python 2.7 or Python 3.4 and up
• requests
• pandas ≥ 0.20
• numpy

$ pip install aiot-studio

Simple datasource usage

from aiotstudio.search import to_pandas
from aiotstudio.logging import log

def execute(parameters):
    log.info("Hello aiot-studio")
             
    result = to_pandas({
        "from": "event",
        "select": [{"count": "*"}]
    })
    return result

Configuration

The library needs to be configured with your mnubo API credentials. The credentials can be shared as follows (by decreasing order of priority):

• Environment variables: MNUBO_CLIENT_ID, MNUBO_CLIENT_SECRET, MNUBO_API_URL
• Local config file: application.conf at the root of the project or anywhere indicated by the MNUBO_CONFIG_FILE environment variable
• Global config file: ~/.settings/mnubo/application.conf (or C:\Users\<username>\.settings\mnubo\application.conf on Windows)

For the last two options, the configuration file should look like this:

[DEFAULT]
mnubo_client_id = {API_CLIENT_ID}
mnubo_client_secret = {API_CLIENT_SECRET}
mnubo_api_url = https://prod.api.mnubo.com

Available package

• search
• blobstore
• logging

Notes:

• All "search" methods expect a MQL query as described in the search API documentation
• Blob store methods are only available when the code runs inside mnubo's architecture, using it locally would throw a FeatureUnavailableError