Installing numpy on Mac to work on AWS Lambda

Question

Is there a way to install numpy on a Mac so that it will work when uploaded to AWS Lambda? I have tried a variety of different ways, including using different pip versions, using easy_install, and following this post, but none of them seem to work. I also tried cloning the git repo and building from there, but I also wasn't able to get that to work (though I'm not sure if I copied the right files up after doing that)

The error I'm getting is:

Unable to import module 'lambda_function': Importing the multiarray numpy extension module failed. Most likely you are trying to import a failed build of numpy. If you're working with a numpy git repo, try git clean -xdf (removes all files not under version control). Otherwise reinstall numpy.

Inspired by this post, I was able to pip install numpy in a Linux environment and get it to work on Lambda.

So my question is: Is it possible to install numpy on a Mac so that it works on AWS Lambda?

Environment: MacBook Pro, MacOS 10.12.2, default python version 2.7.10

I've been testing it with a minor variation on the hello-world-python example on Lambda:

from __future__ import print_function
import numpy

def lambda_handler(event, context):
    #print("Received event: " + json.dumps(event, indent=2))
    print("value1 = " + event['key1'])

(Update) Extending the question: Why do some packages work and others don't?

Answer 1

Update: the preferred approach now is to just use the AWS-provided Lambda Layer for NumPy/SciPy, which is super easy to do.

In the console, select your function and then, under the "Design" section click on "Layers". Then click "Add a Layer", and select "AWSLambda-Python37-SciPy1x" under AWS Provided (or whatever the equivalent is for the version of Python you're using).

Then you can seamlessly import numpy, scipy, etc. into your code with no issues.

10/26/2020 - Added example screenshot: