|
在使用Jupyter练习关联规则挖掘时出现了一些莫名其妙的问题，已解决，记录一下。
给的例子如下：
from efficient_apriori import apriori
import pandas as pd
def data_generator(filename):
"""
Data generator, needs to return a generator to be called several times.
"""
def data_gen():
with open(filename) as file:
for line in file:
yield tuple(k.strip() for k in line.split(','))
#transactions.append(list(line.strip().split(',')))
return data_gen
# file_path = "https://github.com/seratch/apriori.js/blob/master/dataset.csv"
transactions = data_generator("dataset.csv")
itemsets, rules = apriori(transactions, min_support=0.5, min_confidence=1)
itemsets
rules
transactions_2 = data_generator("store_data.csv")
itemsets_2, rules_2 = apriori(transactions_2, min_support=0.0045, min_confidence=0.2)
rules
for rule in rules[:10]:
print(rule)
然后在第四个代码块下报错：
---------------------------------------------------------------------------
TypeError Traceback (most recent call last)
<ipython-input-17-5d285d16ed6b> in <module>()
1 # file_path = "https://github.com/seratch/apriori.js/blob/master/dataset.csv"
2 transactions = data_generator("dataset.csv")
----> 3 itemsets, rules = apriori(transactions, min_support=0.5, min_confidence=1)
C:\ProgramData\Anaconda3\lib\site-packages\efficient_apriori\apriori.py in apriori(transactions, min_support, min_confidence, max_length, verbosity, output_transaction_ids)
61 max_length,
62 verbosity,
---> 63 output_transaction_ids=True,
64 )
65
C:\ProgramData\Anaconda3\lib\site-packages\efficient_apriori\itemsets.py in itemsets_from_transactions(transactions, min_support, max_length, verbosity, output_transaction_ids)
276
277 # Store in transaction manager
--> 278 manager = TransactionManager(transactions)
279
280 # STEP 1 - Generate all large itemsets of size 1
C:\ProgramData\Anaconda3\lib\site-packages\efficient_apriori\itemsets.py in __init__(self, transactions)
29
30 # Populate
---> 31 for i, transaction in enumerate(transactions):
32 for item in transaction:
33 self._indices_by_item[item].add(i)
TypeError: 'function' object is not iterable
查了一下给的例子是官网efficient-apriori的1.0.0以及其他一些版本的?：efficient-apriori · PyPI
最新版本没有这个例子：efficient-apriori · PyPI
研究了好几天，结果发现return那儿加个括号就行，即把return data_gen修改为return data_gen()……
最终稍加修改，包括一些路径和参数，如下：
from efficient_apriori import apriori
import pandas as pd
def data_generator(filename):
"""
Data generator, needs to return a generator to be called several times.
"""
def data_gen():
with open(filename) as file:
for line in file:
yield tuple(k.strip() for k in line.split(','))
#transactions.append(list(line.strip().split(',')))
return data_gen()
#https://pypi.org/project/efficient-apriori/1.0.0/,运行下一句报错，return data_gen修改为return data_gen()解决
# file_path = "https://github.com/seratch/apriori.js/blob/master/dataset.csv"
transactions = data_generator(r"C:\Users\userab\Desktop\第二次实验\dataset.csv")
itemsets, rules = apriori(transactions, min_support=0.1, min_confidence=1)
itemsets
{1: {('Brooklyn',): 216,
('',): 1413,
('MBE',): 953,
('WBE',): 678,
('BLACK',): 427,
('ASIAN',): 287,
('New York',): 419,
('HISPANIC',): 233,
('NON-MINORITY',): 426},
2: {('', 'ASIAN'): 287,
('', 'BLACK'): 423,
('', 'Brooklyn'): 215,
('', 'HISPANIC'): 231,
('', 'MBE'): 946,
('', 'NON-MINORITY'): 426,
('', 'New York'): 418,
('', 'WBE'): 671,
('ASIAN', 'MBE'): 284,
('BLACK', 'MBE'): 427,
('Brooklyn', 'MBE'): 160,
('HISPANIC', 'MBE'): 233,
('MBE', 'New York'): 242,
('MBE', 'WBE'): 240,
('NON-MINORITY', 'New York'): 168,
('NON-MINORITY', 'WBE'): 426,
('New York', 'WBE'): 249},
3: {('', 'ASIAN', 'MBE'): 284,
('', 'BLACK', 'MBE'): 423,
('', 'Brooklyn', 'MBE'): 159,
('', 'HISPANIC', 'MBE'): 231,
('', 'MBE', 'New York'): 241,
('', 'MBE', 'WBE'): 233,
('', 'NON-MINORITY', 'New York'): 168,
('', 'NON-MINORITY', 'WBE'): 426,
('', 'New York', 'WBE'): 248,
('NON-MINORITY', 'New York', 'WBE'): 168},
4: {('', 'NON-MINORITY', 'New York', 'WBE'): 168}}
rules
[{ASIAN} -> {},
{NON-MINORITY} -> {},
{BLACK} -> {MBE},
{HISPANIC} -> {MBE},
{NON-MINORITY} -> {WBE},
{ASIAN, MBE} -> {},
{, BLACK} -> {MBE},
{, HISPANIC} -> {MBE},
{NON-MINORITY, New York} -> {},
{NON-MINORITY, WBE} -> {},
{, NON-MINORITY} -> {WBE},
{NON-MINORITY} -> {, WBE},
{NON-MINORITY, New York} -> {WBE},
{NON-MINORITY, New York, WBE} -> {},
{, NON-MINORITY, New York} -> {WBE},
{NON-MINORITY, New York} -> {, WBE}]
transactions_2 = data_generator(r"C:\Users\userab\Desktop\第二次实验\store_data.csv")
itemsets_2, rules_2 = apriori(transactions_2, min_support=0.0045, min_confidence=0.2)
rules
[{ASIAN} -> {},
{NON-MINORITY} -> {},
{BLACK} -> {MBE},
{HISPANIC} -> {MBE},
{NON-MINORITY} -> {WBE},
{ASIAN, MBE} -> {},
{, BLACK} -> {MBE},
{, HISPANIC} -> {MBE},
{NON-MINORITY, New York} -> {},
{NON-MINORITY, WBE} -> {},
{, NON-MINORITY} -> {WBE},
{NON-MINORITY} -> {, WBE},
{NON-MINORITY, New York} -> {WBE},
{NON-MINORITY, New York, WBE} -> {},
{, NON-MINORITY, New York} -> {WBE},
{NON-MINORITY, New York} -> {, WBE}]
for rule in rules[:10]:
print(rule)
{ASIAN} -> {} (conf: 1.000, supp: 0.202, lift: 1.005, conv: 4929577.465)
{NON-MINORITY} -> {} (conf: 1.000, supp: 0.300, lift: 1.005, conv: 4929577.465)
{BLACK} -> {MBE} (conf: 1.000, supp: 0.301, lift: 1.490, conv: 328873239.437)
{HISPANIC} -> {MBE} (conf: 1.000, supp: 0.164, lift: 1.490, conv: 328873239.437)
{NON-MINORITY} -> {WBE} (conf: 1.000, supp: 0.300, lift: 2.094, conv: 522535211.268)
{ASIAN, MBE} -> {} (conf: 1.000, supp: 0.200, lift: 1.005, conv: 4929577.465)
{, BLACK} -> {MBE} (conf: 1.000, supp: 0.298, lift: 1.490, conv: 328873239.437)
{, HISPANIC} -> {MBE} (conf: 1.000, supp: 0.163, lift: 1.490, conv: 328873239.437)
{NON-MINORITY, New York} -> {} (conf: 1.000, supp: 0.118, lift: 1.005, conv: 4929577.465)
{NON-MINORITY, WBE} -> {} (conf: 1.000, supp: 0.300, lift: 1.005, conv: 4929577.465)
最后吐槽一下，倒数第三个代码块，你这是不是没输出结果啊？
|