Identify a set of L patterns from a binary class dataset
Project description
LPatternIdentification
The formal mathematical definition of the l-Pattern Identification Problem is as follows:
Input:
A finite alphabet Σ, two disjoint sets Good, Bad ⊆ Σn of strings and an integer l > 0
Problem question:
Is there a set of P patterns such that: |P| ≤ l and P → (Good, Bad)?
1. Install:
''' pip install LPatternIdentification '''
2. Load:
''' from LPatternIdentification import feature_set, split_data, get_patterns_from_feature_set, reduce_pattern_set '''
3. Prepare data:
Sort dataset by class labels
Separate observations into numpy ndarray
Separate labels into list
4. Find set of features
''' features = feature_set(observations, labels) '''
5. Split observations by classes
Here, classes are named 'Good' and 'Bad', the 'Good' class being the class of our interest.
''' split_point, Good, Bad = split_data(elections_X, elections_y) '''
6. Return a set of patterns that contain the features and are Good
''' Patterns = get_patterns_from_feature_set(Good, elections_feature_set) '''
7. Identify 'L' number of patterns such that all patterns are uniquely Good and not similar to Bad patterns
''' Patterns_identified = reduce_pattern_set(Patterns, Bad, 7) '''
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 LPatternIdentification-0.0.1.tar.gz
Algorithm | Hash digest | |
---|---|---|
SHA256 | 61390d5acc4f78d2e543368c5e1dd44af257468633e5dab556f795175417f82a |
|
MD5 | be300d0c860c5f07d0a0ac6610de5ff2 |
|
BLAKE2b-256 | 0fd5422afde37fb0cdac433c5ef3538b346a6a528e02ff3ba342aaae421eed70 |
Hashes for LPatternIdentification-0.0.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 6fe68982033b7cdba9403f67903815a525a24c535b927a277c9ed324b5e38351 |
|
MD5 | 71898f27b0131b2358e8e4a253b12c30 |
|
BLAKE2b-256 | 7e67974845e30cc2152cd236319e61038a7e2233f265b6f5f4b7b7b9717d048f |