Import .ipynb files as modules in the system path.
Project description
pip install nostalgiaforever
A Notebook is a Hypothesis
If a notebook is a hypothesis then it can be tested.
Modern scientists are transitioning to a generation where notebooks are common currency. Consider how testing and documentation are crucial aspect of reusable open source software. Those best practices evolved to support the communities developing that software. Generations of open source established idioms and style guides which enable software collaboration at global scale. Now, access to reusable software is blending with modern science, where both disciplines share similar global ambitions to solve increasing complex, multi-objective problems.
Consequently, notebooks have evolved past being a medium for personal insight to become assets for community innovation. They represent extra-personal objects for scientists who have evaluated procedural "units of thought" as computational narratives. In a way, a notebook represents a modern form of hypothesis, and scientists must get comfortable sharing them rapidly. Unfortunately, new authors lack conventions for sharing notebooks as hypotheses within a scientific community. This talk presents tactics from literate programming to create readable, reusable, and reproducible notebooks. These notebook authoring practices promote improved documentation and unit testing.
@docfast presenting behalf of deathbeds
See deathbeds present @ JupyterCon
from nostalgiaforever import importing, testing, reuse
Key Technologies
Key Takeaways
- Notebooks can be used for testing and software.
- Best practices for composing testable computational essays.
- Existing testing frameworks can be used with notebooks.
Order
Before the talk
-
Restart your machine
-
Start the watcher
source activate p6 && cd ahypothesis/nostalgiaforever && watchmedo tricks tricks.yml
-
Prepare the Fira Code fonts https://github.com/deathbeds/jupyterlab-fonts.
-
Hey art nerd! Turn the color on!
-
Cut the release.
Build the docs.
from nbconvert.exporters.markdown import MarkdownExporter
from nbconvert.preprocessors import Preprocessor
class ReplaceLinks(Preprocessor):
def preprocess_cell(self, cell, resources=None, index=0):
if cell['cell_type'] == 'markdown':
if isinstance(cell['source'], list):
cell['source'] = ''.join(cell['source'])
cell['source'] = cell['source'].replace('.ipynb', '')
return cell, resources
from pathlib import Path
if __name__ == '__main__':
from IPython import get_ipython
!jupyter nbconvert --to markdown readme.ipynb
for path in Path('.').rglob('*.ipynb'):
if all(not part.startswith('.') for part in path.parts):
to = ('docs' / path)
print(path)
to.parent.mkdir(exist_ok=True)
MarkdownExporter(preprocessors=[ReplaceLinks()]).from_filename(path)[0]
to.with_suffix('.md').write_text(
("---\n"*2 if len(path.parts) > 1 else "") + MarkdownExporter(preprocess=[ReplaceLinks()]).from_filename(path)[0])
[NbConvertApp] Converting notebook readme.ipynb to markdown
[NbConvertApp] Writing 5882 bytes to readme.md
readme.ipynb
nostalgiaforever/importing.ipynb
nostalgiaforever/plugin.ipynb
nostalgiaforever/reuse.ipynb
nostalgiaforever/testing.ipynb
nostalgiaforever/util.ipynb
test/test_nostalgia.ipynb
Watcher with watchdog
source activate p6 && cd ahypothesis/nostalgiaforever/ && watchmedo tricks tricks.yml
Run unit tests.
if __name__ == '__main__':
try:
__import__('unittest').main('test', argv='--verbose'.split())
except SystemExit: ...
.
----------------------------------------------------------------------
Ran 1 test in 0.001s
OK
# Summary [clear all output then "Run All" -- or it didn't happen]
Summary [clear all output then "Run All" -- or it didn't happen]
if __name__ == '__main__':
!source activate p6 && ipython --profile pidgin -m nostalgiaforever.importing
!source activate p6 && ipython --profile pidgin -m nostalgiaforever.testing
!source activate p6 && ipython --profile pidgin -m nostalgiaforever.reuse
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for nostalgiaforever-0.0.3-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | b47b61169bd65a34c1d30a749f2407692b8a642b3cab199c9c47cc4bdfdd3f2f |
|
MD5 | 7696ce380a1781d4d74718ca55ff4864 |
|
BLAKE2b-256 | a9ed043ce9d9189a6606be0c5747739b3faea93ff475901b10d1bac6cb16308c |