Getting started

[dirknbr]

I am using this for the first time. I wanted to test one of the algos and so ran this. But I get an error, what do I need to put into s?

from timemachines.skaters.simple.thinking import thinking_slow_and_fast
import numpy as np

y = np.arange(100) + np.random.normal(size=100)

res = thinking_slow_and_fast(y, s=None, k=10)

[microprediction]

Hi Dirk,
You feed the skater one data point at a time, and the first time send an empty dictionary.

    y_hat, y_std, s = thinking_slow_and_fast(y[0], s={}, k=10)
    y_hat, y_std, s = thinking_slow_and_fast(y[1], s=s, k=10)
    y_hat, y_std, s = thinking_slow_and_fast(y[2], s=s, k=10)

You can also use the prior or posterior helper functions in skating.py to avoid a loop

Peter