This repo contains Go code for a Lambda function that pulls raw data from the Awair API and saves CSV files to S3.

Create a new Lambda function called awair-to-s3 (if you want to name it something else, you'll have to update main.go). This is how the Lambda function should be configured.

You'll need to configure the following environment variables in Lambda:

• AWAIR_API_KEY After joining the Awair developer program, you can get an access token for pulling data for your device.
• BUCKET The script uses an S3 bucket to store both the compiled function binary and the air quality data. The function code is stored at lambda/awair-to-s3.zip, and the CSV files are stored at airdata/[TIMESTAMP].csv.
• DEVICE_TYPE Both device type and ID can be found by listing out devices through the API.
• DEVICE_ID

Once you have the function setup, you'll want to create a trigger for it to run. To run the function every hour, create a new CloudWatch Events trigger with a Schedule Expression equal to rate(1 hour). If you want to change the frequency of how often the Lambda function is called, just change the rate in the Schedule Expression. For more on CloudWatch Schedule Expressions, see this documentation.

After configuring the function in Lambda, you can use deploy.sh to compile the code, upload it to S3, and update the function to use it. The BUCKET variable should be updated. You'll need AWS command line tools and Go installed.

The function runs once per hour and saves a new CSV file with three columns: timestamp, sensor, and value. Here's what that looks like:

2020-01-01 16:59:55.817000,temp,23.809999465942383
2020-01-01 16:59:55.817000,humid,32.84000015258789
2020-01-01 16:59:55.817000,co2,829
2020-01-01 16:59:55.817000,voc,137
2020-01-01 16:59:55.817000,pm25,3

With this date format Athena can query this directly from S3.