CODEWITHSUNDEEP
pythonapriori
DavidM
DavidM

Reputation: 57

Python Apyori Sorting by Lift

So I've been using Apyori to get the minimum support from a dataset of Titanic survivors.

rules = apriori(titanic, min_support = 0.1, min_confidence = 1.0)
print(list(rules))

Here's an example of one of the elements I'd get as output

[RelationRecord(items=frozenset({'Crew', 'Adult'}), 
support=0.4020899591094957, ordered_statistics= 
[OrderedStatistic(items_base=frozenset({'Crew'}), 
items_add=frozenset({'Adult'}), confidence=1.0, lift=1.0521032504780115)])

However, I'm trying to sort by lift, and I'm not sure how to approach this problem, since the lift element seems to be inside a tuple, but I'm not sure exactly on how I'd go about sorting it.

Any help is greatly appreciated.

Thanks.

Upvotes: 3

Views: 3097

Answers (2)

ShamusTinker
ShamusTinker

Reputation: 1

Here's an easier method I found while looking for the exact same answer.

rules = apriori(titanic, min_support = 0.1, min_confidence = 1.0)

rules = rules.sort_values(['confidence', 'support'], ascending =[False, False]) 

print(list(rules))

Upvotes: 0

Okroshiashvili
Okroshiashvili

Reputation: 4149

You can put your output in Pandas DataFrame and then sort by lift

You have this:

rules = apriori(titanic, min_support = 0.1, min_confidence = 1.0)

And let create results

results = list(rules)

Now we want to write our result into Pandas DataFrame and then sort by Lift

df = pd.DataFrame(columns=('Items','Antecedent','Consequent','Support','Confidence','Lift'))

Support =[]
Confidence = []
Lift = []
Items = []
Antecedent = []
Consequent=[]

for RelationRecord in results:
    for ordered_stat in RelationRecord.ordered_statistics:
        Support.append(RelationRecord.support)
        Items.append(RelationRecord.items)
        Antecedent.append(ordered_stat.items_base)
        Consequent.append(ordered_stat.items_add)
        Confidence.append(ordered_stat.confidence)
        Lift.append(ordered_stat.lift)

df['Items'] = list(map(set, Items))                                   
df['Antecedent'] = list(map(set, Antecedent))
df['Consequent'] = list(map(set, Consequent))
df['Support'] = Support
df['Confidence'] = Confidence
df['Lift']= Lift

Sort resulted dataframe by Lift

df.sort_values(by ='Lift', ascending = False, inplace = True)

Upvotes: 12

Related Questions