Skip to main content

​
Introduction

This document provides detailed guidelines on using the integrated Python editor console within our platform. The editor is pre-configured with essential libraries and functionalities to assist in executing Python scripts effectively, particularly for data querying and handling secrets.

​
Where to find the Python Editor Console?

  • In the Create Metric page, click on the “Custom Query” option located at the top left side of the page.
  • Then, choose </> Python from the dropdown in the </> Query Editor panel to access our Python Editor Console.
Python Edit Console

​
Pre-installed Libraries

  • Requests Module: The requests library is readily available and pre-imported for HTTP requests.
    You do not need to import it again in your scripts.
import requests  # This module is already available in the scope

​
Working with Variables

To integrate dynamic data within your scripts, you can utilize predefined variables and data structures:
  • Client ID: Use the variable client_id which should be a string representing the client identifier.
  • Metric Filters: This variable metric_filters allows for the filtering of data based on various criteria. The structure of metric_filters is defined as follows:
from typing import Dict, Optional, Union
import datetime

class DateDict:
    startDate: datetime.date
    endDate: datetime.date

# MetricFilters type annotation
MetricFilters = Dict[str, Union[DateDict, str, int]]
The metric_filters variable can include filters of types STRING, DATE, and NUMBER, uniquely identified by a filter name:
  • STRING: A plain string value.
  • DATE: A dictionary containing startDateand endDate, both of datetime.date type.
  • NUMBER: An integer value.
When is metric_filters populated?
metric_filters is populated only when (1) the metric’s SQL query actually uses that filter variable (e.g. in a WHERE clause), and (2) the metric is run in a dashboard context where that filter is applied and the user has selected a value. If you see empty {} even when a filter is selected, check that the metric’s SQL references the filter variable by the exact name, and that you are running from a dashboard view with the filter applied—not from a standalone console or a metric that does not use the filter.

​
Handling Secrets

Secrets are stored and accessed via the secrets dictionary. This allows secure storage and retrieval of sensitive data such as API keys or database credentials. You can set up your secrets in the Home Page → Settings Tab → Secrets. 
url = secrets['BASE_URL']  # Accessing the base URL stored in secrets
res = requests.get(url).json()  # Making a GET request using the retrieved URL

​
Storing Results

To store or output data from your script, assign your data to the resultvariable. This variable should be an array of objects (or a list of dictionaries in Python terms): 
result = [{'key': 'value'}, ...]  # An example assignment to the result variable
result = requests.get('SOME URL').json()  # Storing JSON response from a GET request

​
Usage Example

Here’s are some quick examples demonstrating how to utilize these capabilities: 
country_population = {
    'USA': 331000000,
    'India': 1380000000,
    'China': 1430000000,
    'Brazil': 212600000,
    'Nigeria': 206000000
}

result = [{'country': country, 'population': population} for country, population in country_population.items()]

# Assuming client_id and metric_filters are predefined
client_id = '12345'
metric_filters = {'salesDate': {'startDate': datetime.date(2021, 1, 1), 'endDate': datetime.date(2021, 12, 31)}}

# Accessing a secret
api_url = secrets['API_URL']

# Making a GET request
response = requests.get(f"{api_url}/data?client_id={client_id}").json()

# Assigning response to result
result = response
This guide should help you effectively utilize the Python editor console for your data handling needs. The capabilities outlined here are designed to enhance the security, flexibility, and efficiency of your data manipulation tasks.