Mercurial > repos > bgruening > sklearn_regression_metrics
annotate association_rules.py @ 29:7c04f3c9f674 draft
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
| author | bgruening | 
|---|---|
| date | Fri, 30 Apr 2021 23:46:14 +0000 | 
| parents | |
| children | e3e0ae93f7c8 | 
| rev | line source | 
|---|---|
| 29 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 1 import argparse | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 2 import json | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 3 import warnings | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 4 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 5 import pandas as pd | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 6 from mlxtend.frequent_patterns import association_rules, fpgrowth | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 7 from mlxtend.preprocessing import TransactionEncoder | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 8 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 9 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 10 def main(inputs, infile, outfile, min_support=0.5, min_confidence=0.5, min_lift=1.0, min_conviction=1.0, max_length=None): | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 11 """ | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 12 Parameter | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 13 --------- | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 14 input : str | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 15 File path to galaxy tool parameter | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 16 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 17 infile : str | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 18 File paths of input vector | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 19 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 20 outfile : str | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 21 File path to output matrix | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 22 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 23 min_support: float | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 24 Minimum support | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 25 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 26 min_confidence: float | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 27 Minimum confidence | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 28 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 29 min_lift: float | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 30 Minimum lift | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 31 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 32 min_conviction: float | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 33 Minimum conviction | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 34 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 35 max_length: int | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 36 Maximum length | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 37 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 38 """ | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 39 warnings.simplefilter('ignore') | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 40 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 41 with open(inputs, 'r') as param_handler: | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 42 params = json.load(param_handler) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 43 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 44 input_header = params['header0'] | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 45 header = 'infer' if input_header else None | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 46 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 47 with open(infile) as fp: | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 48 lines = fp.read().splitlines() | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 49 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 50 if header is not None: | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 51 lines = lines[1:] | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 52 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 53 dataset = [] | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 54 for line in lines: | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 55 line_items = line.split("\t") | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 56 dataset.append(line_items) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 57 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 58 # TransactionEncoder learns the unique labels in the dataset and transforms the | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 59 # input dataset (a Python list of lists) into a one-hot encoded NumPy boolean array | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 60 te = TransactionEncoder() | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 61 te_ary = te.fit_transform(dataset) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 62 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 63 # Turn the encoded NumPy array into a DataFrame | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 64 df = pd.DataFrame(te_ary, columns=te.columns_) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 65 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 66 # Extract frequent itemsets for association rule mining | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 67 # use_colnames: Use DataFrames' column names in the returned DataFrame instead of column indices | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 68 frequent_itemsets = fpgrowth(df, min_support=min_support, use_colnames=True, max_len=max_length) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 69 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 70 # Get association rules, with confidence larger than min_confidence | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 71 rules = association_rules(frequent_itemsets, metric="confidence", min_threshold=min_confidence) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 72 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 73 # Filter association rules, keeping rules with lift and conviction larger than min_liftand and min_conviction | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 74 rules = rules[(rules['lift'] >= min_lift) & (rules['conviction'] >= min_conviction)] | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 75 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 76 # Convert columns from frozenset to list (more readable) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 77 rules['antecedents'] = rules['antecedents'].apply(list) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 78 rules['consequents'] = rules['consequents'].apply(list) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 79 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 80 # The next 3 steps are intended to fix the order of the association | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 81 # rules generated, so tests that rely on diff'ing a desired output | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 82 # with an expected output can pass | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 83 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 84 # 1) Sort entry in every row/column for columns 'antecedents' and 'consequents' | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 85 rules['antecedents'] = rules['antecedents'].apply(lambda row: sorted(row)) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 86 rules['consequents'] = rules['consequents'].apply(lambda row: sorted(row)) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 87 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 88 # 2) Create two temporary string columns to sort on | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 89 rules['ant_str'] = rules['antecedents'].apply(lambda row: " ".join(row)) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 90 rules['con_str'] = rules['consequents'].apply(lambda row: " ".join(row)) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 91 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 92 # 3) Sort results so they are re-producable | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 93 rules.sort_values(by=['ant_str', 'con_str'], inplace=True) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 94 del rules['ant_str'] | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 95 del rules['con_str'] | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 96 rules.reset_index(drop=True, inplace=True) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 97 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 98 # Write association rules and metrics to file | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 99 rules.to_csv(outfile, sep="\t", index=False) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 100 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 101 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 102 if __name__ == '__main__': | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 103 aparser = argparse.ArgumentParser() | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 104 aparser.add_argument("-i", "--inputs", dest="inputs", required=True) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 105 aparser.add_argument("-y", "--infile", dest="infile", required=True) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 106 aparser.add_argument("-o", "--outfile", dest="outfile", required=True) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 107 aparser.add_argument("-s", "--support", dest="support", default=0.5) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 108 aparser.add_argument("-c", "--confidence", dest="confidence", default=0.5) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 109 aparser.add_argument("-l", "--lift", dest="lift", default=1.0) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 110 aparser.add_argument("-v", "--conviction", dest="conviction", default=1.0) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 111 aparser.add_argument("-t", "--length", dest="length", default=5) | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 112 args = aparser.parse_args() | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 113 | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 114 main(args.inputs, args.infile, args.outfile, | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 115 min_support=float(args.support), min_confidence=float(args.confidence), | 
| 
7c04f3c9f674
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit ea12f973df4b97a2691d9e4ce6bf6fae59d57717"
 bgruening parents: diff
changeset | 116 min_lift=float(args.lift), min_conviction=float(args.conviction), max_length=int(args.length)) | 
