ericmjl

from PIL import Image
import numpy as np
from pathlib import Path

def open(fname):
    img = Image.open(fname).resize((384, 640))
    return img


def array(img):

ericmjl

# This script is intended for downloading all four videos on causality by
# Jonas Peters (University of Copenhagen) from YouTube.
#
# Copyright belongs to their appropriate copyright holders. I am only
# providing this script for convenience.
#
# To run this script, first ensure that you have the Python package
# `youtube-dl` installed. Assuming you are able to install it into your
# favourite computing environment, it should be a single command:
#

ericmjl

name: pymc4-dev
channels:
- defaults
- conda-forge
- ericmjl
dependencies:
- python=3.6
- jupyter
- jupyterlab
- conda

ericmjl

ericmjl

============================= test session starts ==============================
platform darwin -- Python 3.6.7, pytest-4.0.2, py-1.7.0, pluggy-0.8.0
rootdir: /Users/ericmjl/github/software/autograd-sparse, inifile:
collected 3 items / 2 deselected

tests/test_sparse.py F                                                   [100%]

=================================== FAILURES ===================================
_____________________________ test_sparse_dot_grad _____________________________

ericmjl

ericmjl

ericmjl

name: dl-workshop
channels:
- defaults
- conda-forge
- ericmjl
dependencies:
- python=3.7
- jupyter
- jupyterlab
- conda

ericmjl

ericmjl

from sklearn.metrics import explained_variance_score

def var_explained(preds, actual):
    """
    Implementation taken directly from the formula on this page:
    http://scikit-learn.org/stable/modules/model_evaluation.html#explained-variance-score
    """
    return 1 - ((preds - actual).var() / actual.var())

y_pred = np.array([3, -0.5, 2, 7])