Skip to content

Latest commit

 

History

History
56 lines (34 loc) · 4.72 KB

CONTRIBUTE_BATCH_STYLE_MODELS.md

File metadata and controls

56 lines (34 loc) · 4.72 KB
Raw

Contribution pattern: new batch-style time-series package

Do you have a favourite Python package for time-series analysis that you would like to see included in the funnel? There's some discussion on this LI post

When to use this pattern.

  • You maintain a time-series package, or just really like one.
  • The package exposes a way of predicting time-series in an offline (sometimes tabular) fashion (as compared with online, incremental)

Just to emphasize, if the package is, say a state-space model, or some other incremental method, then it would be wasteful to follow the pattern outlined here. See CONTRIBUTE_ONLINE_STYLE_MODELS.md instead.

How to contribute

  1. Look at the example provided by sk skaters directory and try to grok
  2. If step (1) fails, ask questions on slack (invite here)
  3. Grok the package you think should be in. Create an example colab notebook (like examples here) that uses the package. It should show how to produce a k-vector of 1..k step ahead predictions. You'd be surprised at how many packages seem to think this is an obscure use case and don't include it in their README :)

At this point you've already helped a lot. If you want to take it all the way...

  1. Choose a short PREFIX that isn't exactly the same as the library (here PREFIX='sk', obviously)
  2. Write PREFIXinclusion.py (in the same style as skinclusion.py)
  3. Write PREFIXwrappers.py (expose the batch functionality in the same style as skwrappers.py)

Thus far this has exposed the 3rd party library in an offline manner (i.e. expects fit to be called each data point). Now you need to turn this into a skater.

  1. Read or re-read the main README.md that explains the skater signature
  2. Read the short pithy example skautoarima and notice how it makes use of batchskater.py that does all the work for you.
  3. Create one or more skater files similar to skautoarima
  4. For the convenience of others, adopt the style of putting a mini-test at the bottom, as with skautoarima.

Next you need to register a list of fully autonomous skaters

  1. Create a file like allskskaters

Write unit tests

  1. Cut and paste the example unit test test_sk_random_skater. Modify it.

Run pytest and the individual test repeatedly. I rely on Github actions to pytest package combinations, but I suppose you can use act to run the actions locally. When it is working well you can include it in the master list:

  1. Modify localskaters.py
  2. Modify README.md to include a suggestion to users to pip install the new requirements
  3. Modify setup.py to include the new directory.

More help / discussion

As noted, I try to jump on a Google Meet twice a week and the details are in the microprediction knowledge center. My arrival rate is higher on Fridays than Tuesdays :)

I'm not so good at scheduling calls outside of these times and frankly that tends to be counter to my productivity anyway. So just jump on some Tuesday night or Friday noon if you are keen to contribute to this package, or anything else that relates to open source community prediction.