A flexible pairwise tool written in Python.
Project description
Requirements
Python: 3.3 or later.
Tested with 3.7, 3.11
Installation
$ pip install covertable
Usage
Just import covertable and call make function.
>>> from covertable import make, sorters, criteria
>>> machine_list = ['iphone', 'pixel']
>>> os_list = ['ios', 'android']
>>> browser_list = ['FireFox', 'Chrome', 'Safari']
>>> # list input and output
>>> make(
... [machine_list, os_list, browser_list], # list factors
... length=2, # default: 2
... sorter=sorters.random, # default: sorters.hash
... criterion=criteria.simple, # default: criteria.greedy
... seed=100, # default: ''
... pre_filter=lambda row: not(row[1] == 'android' and row[0] != 'pixel') and not(row[1] == 'ios' and row[0] != 'iphone'), # default: None
... )
[
['pixel', 'android', 'Safari'],
['iphone', 'ios', 'Chrome'],
['iphone', 'ios', 'Safari'],
['pixel', 'android', 'Chrome'],
['pixel', 'android', 'FireFox'],
['iphone', 'ios', 'FireFox']
]
>>> # dict input and output
>>> make(
... {'machine': machine_list, 'os': os_list, 'browser': browser_list}, # dict factors
... length=2, # default: 2
... tolerance=3, # default: 0
... post_filter=lambda row: not(row['os'] == 'android' and row['machine'] != 'pixel') and not(row['os'] == 'ios' and row['machine'] != 'iphone'), # default: None
... )
[
{'machine': 'pixel', 'browser': 'Chrome', 'os': 'android'},
{'machine': 'pixel', 'browser': 'FireFox', 'os': 'android'},
{'machine': 'iphone', 'os': 'ios', 'browser': 'Chrome'},
{'os': 'ios', 'browser': 'FireFox', 'machine': 'iphone'}
]
Options
covertable.make function has options as keyword argument.
All options are omittable.
length
Number of factors to be covered. (default: 2)
Obviously the more it increases, the more number of combinations increases.
sorter
Combinations depend on the order of spreading all over the rows.
You can choice a sorter from the following:
- sorters.random:
This makes different combinations everytime. (fastest)
- sorters.hash:
This makes combinations depending on hash of the pair and seed. (default)
It receives seed and useCache options.
seed option decides the order of storing from unstored pairs, therefore it outputs the same result every time when number of factors and seed are the same.
useCache option decide if using cache of hash or not. (default: true)
It is around 10% faster than setting useCache off.
criterion
- criteria.simple:
This extracts any pairs that can be stored into the processing row.
- criteria.greedy:
This attempts to make most efficient combinations. (default)
These combinations are not always shorter than simple criterion.
It receives tolerance option.
pre_filter
This means a function to filter beforehand.
It receives an argument row as object type.
When the function returns False, the row combination will not be registered.
If factors type is Array, you should specify an index at the subscript like row => row[1] < 6.
If factors type is Object, you should specify a key at the subscript like row => row['month'] < 6
post_filter
This means a function to filter later.
The usage is the same as preFilter, only the difference is the timing of the call. It will delete rows not matched this function at the last.
For this reason, the final test cases may not satisfy the factors coverage.
Development
# preparation
$ python3 -m venv venv
$ source venv/bin/activate
(venv) $ pip install -r dev_requirements.txt
# testing
(venv) $ pytest
Publish
(venv) $ python setup.py sdist bdist_wheel
(venv) $ twine upload --repository pypi dist/*
More info
Project details
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
File details
Details for the file covertable-2.1.0.tar.gz
.
File metadata
- Download URL: covertable-2.1.0.tar.gz
- Upload date:
- Size: 7.9 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.10.10
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | e123e2206fc02b8a7b997ec8922f13d8fa1ba436ff5e8de522a5fa9fc72f4d41 |
|
MD5 | bfb6322b8c8836f6c5aa264e1a4afda9 |
|
BLAKE2b-256 | 4819b3a67b537c66ce07a38d1038ee34f851a583847ec8a24dbb9a9aa9a0eebd |
File details
Details for the file covertable-2.1.0-py3-none-any.whl
.
File metadata
- Download URL: covertable-2.1.0-py3-none-any.whl
- Upload date:
- Size: 7.1 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.10.10
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 44c986d000ea3c41789e876da29ac94d091a275c78b3a110edbdb38768631fe7 |
|
MD5 | d9ea270265684a198afb4d1b63874d9d |
|
BLAKE2b-256 | cb9320f3fb904ee02daad036b3d2cee6cc1185755c27817aa4b39ec8fe10a3ea |