misc.md

Miscellaneous

Benchmarking

1. Custom timer

import time
get_time = time.perf_counter
t0 = get_time()
### some code ###
t1 = get_time()
print('Time elapsed = {}'.format(t1 - t0))

2. Timeit module (example taken from here)

def my_function():
    try:
        1 / 0
    except ZeroDivisionError:
        pass
 
if __name__ == "__main__":
    import timeit
    setup = "from __main__ import my_function"
    print timeit.timeit("my_function()", setup=setup)

Path names

import os
pathname = os.path.join('dir1', 'dir2', 'filename')
# MacOS/Linux: dir1/dir2/filename
# Windows: dir1\dir2\filename

Executing in terminal

To run script.py you can use the following terminal command:

python script.py

However, if you want to run it as an executable file without calling the python command, just add the following text to the head of script.py:

#!/usr/bin/env python
# coding: utf-8

Then you can run the script using

./script.py

Packages

Useful resources:

• https://packaging.python.org/tutorials/packaging-projects/
• https://kiwidamien.github.io/making-a-python-package.html
• https://medium.com/@joel.barmettler/how-to-upload-your-python-package-to-pypi-65edc5fe9c56

Load package from inside directory:

python -m pip install .

Reload package (e.g., if made edits):

python -m pip install --upgrade --force-reinstall ..

Save/load data (json)

Note that this does not work for np.ndarrays - for those see NumPy section or pickle (below).

import json
data = {'a' : [1, 2, 3], 'b' : ('hello', 'world')} # dictionary

# save
with open('data.json', 'w') as outfile:
  json.dump(data, outfile)
  
# load
with open('data.json', 'r') as infile:
  data = json.load(infile)

Save/load data (pickle)

Can save dictionary of np.ndarrays using pickle.

import pickle
pickle.dump(dict_name, open("save.p", "wb"))
dict_name = pickle.load(open("save.p", "rb"))

Find n smallest/largest elements

Use heaps:

import heapq
a = [4, 6, 2, 3, 7]
heapq.nsmallest(3, a) # [2, 3, 4]
heapq.nlargest(3, a) # [7, 6, 4]

PEP8 coding style

To check code against the PEP8 style guidelines can use the pycodestyle or flake8 modules (see links for installation). More notes on PEP8 here.

Check python script:

pycodestyle --first script.py

Check Jupyter notebook:

1. Install pycodestyle_magic

pip install flake8 pycodestyle_magic

2. Inside notebook cell (see pycodestyle_magic notes):

%load_ext pycodestyle_magic
%pycodestyle_on
%pycodestyle_off

Use black to reformat python file to match PEP8:

pip install black
black messy_file.py