EikonAI

On demand location intelligence.

Eikonai is a lightweight Python SDK for the EIKON Location APIs: compare places, find places that contain X, and monitor change over time—without the hassle of GIS data wrangling.

• Context - "What does location X contain?"
• Similarity — “How similar is location X and location Y?”
• Change — “How much did this place change since then?”
• Models - Some models to use on demand for object detection and other helpful tasks

Works nicely in data notebooks, Streamlit apps, and backend services.

Install (dev)

pip install eikonsai

Python: 3.9–3.12 OS: macOS, Linux, Windows

Usage

Quick start

from eikonsai import utils, context, similarity, models

# set your api_key 
import os


# 1. Use the utils module to get your api_key
my_api_key = utils.get_api_key_from_credentials(email="myregisteredemail",
                                                password="mypassword")
os.environ["user_api_key"] = my_api_key

# 2. Get a location description for a given place 

location_1 = [51.531143, -0.159893] # regent's park
location_1_description = context.get_location_description(lat = location_1[0],
                                lon = location_1[1], 
                                resolution="low",
                                user_api_key=os.environ["user_api_key"]
                                )
print(location_1_description)


# 3. Comparing the visual similarity of two different locations

location_1 = [51.531143, -0.159893] # regent's park
location_2 = [51.433727, -0.214443] # wimbledon tennis club
resolution = "high"
visual_similarity = similarity.visual_similarity(location_1_lat_lon_list=location_1,
                             location_2_lat_lon_list=location_2, 
                             resolution=resolution,
                             user_api_key=os.environ["user_api_key"]
                            )
print(visual_similarity)

The official documentation and more extensive examples for Eikon can be found here

Background

All qualifying education email addresses will be provided with free credits for each month for research purposes. To access these credits you must be enrolled on a UK recognised postgraduate degree programme listed here.

Update for version 0.9.3

Added a utils function to allow users to check which areas Eikon currently supports for filtering with Search API.

Update for version 0.9.2

An additional function has been added to the utils package to allow retrieval of previous search results from the search_api.

Update for version 0.9.1

Version 0.9.1+ brings one major new module called jobs. This is a module that allows for asynchronous processing of user queries.

• Search - the search_api is for specific search focused queries such as " I am looking to find a solar panel farm in an agricultural area that is accessible only via small roads". This function has the added functionality of being able to specify named areas that you would like to focus your search queries in (for example i.e. Camden, Hillingdon etc.). This function runs in an asynchronous manner meaning that requests will be processed based on compute availability. An example of how to run this function is provided on the Eikon SDK Github page.

• Porfolio comparison - the portfoilio_comparison function allows for comparisons based on similarity of two lists of location pairs in an efficient manner. For example it is possible to compare Wembley stadium to all locations in Greater London for similarity within ~5 mins. An example of how to run this function is provided on the Eikon SDK Github page.

We hope you will enjoy using the jobs module. This represents a new approach to using machine learning and AI together with geospatial processing techniques to achieve things that weren't possible a few years ago.