Measure Lambda function

This AWS Lambda function is designed to predict the size of parts of body based on height and weight.

If you want to develop the machine learning model and how to deploy the function to AWS, check Deployment section

If you just want to use the API, check Endpoint section

Deployment using AWS CDK and Python Flask

Prerequisites

Before deploying the Python API Service, ensure you have the following:

AWS CLI installed and configured with the necessary permissions (run aws configure).
Docker installed on your local machine.
AWS CDK installed (npm install -g aws-cdk).

Deployment Steps

Clone the Repository to local machine or server:

git clone https://github.com/your-repo/your-python-api.git
cd measureup-docker-aws

Develop Python machine learning pipeline

To modify the Python code which is used for prediction, check the image/ folder and you can find the lambda_function.py which handles the main ML project.

The models for predicting measurements are stored in the image/src/male and image/src/female folders.

These models are used in the api_test.py script and you can feel free to modify the script or add your own models.

Build the Docker Image:

The image/Dockerfile includes the instructions for turning the ML app into a Docker image. In the image/ folder, you can run:

docker build -t  measureup-model:test .

To run the Docker image locally, run:

docker run -p 9090:8080 measureup-model:test

Create AWS infrascture

To configure the AWS lambda function, go to lib/ and edit the measureup_aws-stack.ts file. In this file you can find how the Docker image function is defined.

Deploy using AWS CDK:

Before you deploy, please make sure you have your AWS CLI configured, since you will need that to interact with AWS with your account. To perform this step, you can run aws sts get-caller-identity, and you are supposed to get response that includes details about your AWS account.

And make sure you run the boostrap command: cdk bootstrap --region [your_region]. Learn more about Bootstrapping

Since we are using Typescript to work with AWS CDK, make sure you run npm install to install npm packages.

Now you can just run to deploy the Lambda function.

cdk deploy

It is likely to take several minutes to run this command, after that you will find the URL for the function in the console.

API USAGE GUIDE

To use this API, you can follow the instructions below. Make sure you obtain a valid api key from the project maintainer.

Base URL: https://[endpoint].[region].amazonaws.com/v1

URL: /measure

Method: POST

Authentication: All requests to the API must include the x-api-key header with a valid API key.

Parameters:

Name	Type	Description	Allowed values
`gender`	`string`	gender of the target t he individual	`["male", "female"]`
`part`	`string`	indicate which part of the body needs to be predicted	`["chest", "waist", "hip", "thigh", "ankle", "arm", "outseam",]`
`weight`	`float`	weight of the individual in kilogram
`height`	`float`	height of the individual in centimeter

The Lambda function is triggered by a URL with the following structure:

https://[endpoint].on.aws/?height=[height]&weight=[weight]&gender=[gender]&part=[part]

Note: Parameters are case-sensitive. Ensure correct casing when providing values for parameters.

Example :

curl -X POST -H "x-api-key: ${INSERT_YOUR_KEY}" "https://[endpoint].[region].amazonaws.com/v1/measure?height=179&weight=84&gender=male&part=waist"

curl -X POST -H "x-api-key: ${INSERT_YOUR_KEY}" "https://[endpoint].[region].amazonaws.com/v1/measure?height=155&weight=59&gender=female&part=chest"

curl -X POST -H "x-api-key: ${INSERT_YOUR_KEY}" "https://[endpoint].[region].amazonaws.com/v1/measure?height=161.9&weight=62.3&gender=female&part=chest"

Returned Object

The Lambda function returns an object with the following structure: