def fit(self, rule_cutoff):
dataframes = self._prepare_dataframes()
scores = []
for dataframe_train, dataframe_test in dataframes:
txns_train = TransactionDB.from_DataFrame(dataframe_train)
rules = top_rules(txns_train.string_representation, appearance=txns_train.appeardict)
cars = createCARs(rules)[:rule_cutoff]
quant_dataframe_train = QuantitativeDataFrame(dataframe_train)
quant_dataframe_test = QuantitativeDataFrame(dataframe_test)
self.classifier.fit(quant_dataframe_train, cars, debug=self.debug)
score = None
if self.score_auc:
score = self.classifier.score_auc(quant_dataframe_test)
else:
score = self.classifier.score(quant_dataframe_test)
scores.append(score)
return scores
def fit(self,
quant_dataframe,
cars=None,
rule_cutoff=30,
lambda_array=7 * [1],
class_name=None,
debug=False,
algorithm="SLS"):
self.quant_dataframe_train = quant_dataframe
self._prepare(quant_dataframe, class_name)
for class_, clf_dict in self.ids_classifiers.items():
print("training class:", class_)
clf = clf_dict["clf"]
quant_dataframe = clf_dict["quant_dataframe"]
pandas_dataframe = quant_dataframe.dataframe
txns = TransactionDB.from_DataFrame(pandas_dataframe)
rules = top_rules(txns.string_representation,
appearance=txns.appeardict)
cars = createCARs(rules)
cars.sort(reverse=True)
clf.fit(quant_dataframe,
cars[:rule_cutoff],
lambda_array=lambda_array,
debug=debug,
algorithm=algorithm)
def mine_CARs(df, rule_cutoff, sample=False):
txns = TransactionDB.from_DataFrame(df)
rules = top_rules(txns.string_representation, appearance=txns.appeardict)
cars = createCARs(rules)
cars_subset = cars[:rule_cutoff]
if sample:
cars_subset = random.sample(cars, rule_cutoff)
return cars_subset
def mine_rules(self, pandas_df, minsup=0.2):
frequent_itemsets = self.mine_frequent_itemsets(pandas_df, minsup)
distinct_classes = list(pandas_df.iloc[:, -1].unique())
fim_rules = self._convert_to_fim_rules(frequent_itemsets,
distinct_classes,
pandas_df.columns[-1])
cars = createCARs(fim_rules)
print(cars)
for car in cars:
car.confidence = self._calculate_rule_confidence(car, pandas_df)
return cars
def test_createCARs(self):
generated_rules = [('Y:=:1', (), 0.5, 0.5), ('Y:=:0', (), 0.5, 0.5),
('Y:=:1', ('A:=:1', ), 0.5, 1 / 3)]
cars = createCARs(generated_rules)
assert cars[0].consequent == Consequent("Y", 1)
assert cars[0].confidence == 0.5
assert cars[0].support == 0.5
assert cars[1].consequent == Consequent("Y", 0)
assert cars[1].confidence == 0.5
assert cars[1].support == 0.5
assert cars[2].consequent == Consequent("Y", 1)
assert cars[2].antecedent == Antecedent([Item("A", 1)])
assert cars[2].confidence == 1 / 3
assert cars[2].support == 0.5
def mine_CARs(df,
rule_cutoff,
sample=False,
random_seed=None,
**top_rules_kwargs):
if random_seed:
random.seed(random_seed)
np.random.seed(random_seed)
txns = TransactionDB.from_DataFrame(df)
rules = top_rules(txns.string_representation,
appearance=txns.appeardict,
**top_rules_kwargs)
cars = createCARs(rules)
cars_subset = cars[:rule_cutoff]
if sample:
cars_subset = random.sample(cars, rule_cutoff)
return cars_subset
import pandas as pd
import numpy as np
from pyarc.data_structures import TransactionDB
from pyarc.algorithms import top_rules, createCARs
df = pd.read_csv("./data/iris0.csv")
df[df["sepallength"] == "-inf_to_5.55"] = np.NaN
print(df)
txns = TransactionDB.from_DataFrame(df)
rules = top_rules(txns.string_representation, appearance=txns.appeardict)
cars = createCARs(rules)
for car in cars[:10]:
print(car)
