Extract string from rules frozensets

Question

With the following statement:

rules = association_rules(frequent_itemsets, metric="lift", min_threshold=1.2)

I get a data frame of rules in the format:

frozenset({'Co_Apples'})

But I need to extract a Co_Apples as a string.

How can I do that?

You seem to have a set of one element. You can convert it to a list, and take the first element: `list(rules)[0]`. — 9000, Sep 24 '18 at 18:26
thanks @9000, the one item list is just an example but converting it to a list may be a good idea. I'll try — Erez Ben-Moshe, Sep 25 '18 at 21:07

score 8 · Answer 1 · answered Nov 08 '19 at 03:07

8

rules["antecedents"] = rules["antecedents"].apply(lambda x: ', '.join(list(x))).astype("unicode")

It is work for me. Thanks Frank Herfert save my day!

answered Nov 08 '19 at 03:07

Johnny Hsiao

81
1
2

score 5 · Accepted Answer · answered Mar 19 '19 at 21:08

5

You can use the following code to get a string from frozenset type columns and then cast the string to unicode.

rules["antecedents"] = rules["antecedents"].apply(lambda x: list(x)[0]).astype("unicode")
rules["consequents"] = rules["consequents"].apply(lambda x: list(x)[0]).astype("unicode")

answered Mar 19 '19 at 21:08

Tomasz Cieplak

66
1
3

5

In case you have item combinations (a frozenset with more than one value), using `list(x)[0]` will only show the first value. You can use `', '.join( list(x) )` instead to separate the items with a comma. `rules["antecedents"] = rules["antecedents"].apply(lambda x: ', '.join(list(x))).astype("unicode")` – Frank Herfert Jul 18 '19 at 15:25

Extract string from rules frozensets

2 Answers2