Collaborative filtering was updated to avoid problems caused by recen…

[SErfanR]

درود استاد گرامی

بخش هایی از کد collaborative filtering بدلیل تغییرات صورت گرفته در pandas و سایر کتابخانه ها، در نسخه های جدید امکان اجرا نداشت.
تغییرات کوچکی صورت گرفته تا از این مشکلات جلوگیری شود.

1.

old:

#Using regular expressions to find a year stored between parentheses
#We specify the parantheses so we don't conflict with movies that have years in their titles
movies_df['year'] = movies_df.title.str.extract('(\(\d\d\d\d\))',expand=False)
#Removing the parentheses
movies_df['year'] = movies_df.year.str.extract('(\d\d\d\d)',expand=False)
#Removing the years from the 'title' column
movies_df['title'] = movies_df.title.str.replace('(\(\d\d\d\d\))', '', regex=True)
#Applying the strip function to get rid of any ending whitespace characters that may have appeared
movies_df['title'] = movies_df['title'].apply(lambda x: x.strip())

new:

#Using regular expressions to find a year stored between parentheses
#We specify the parantheses so we don't conflict with movies that have years in their titles
movies_df['year'] = movies_df.title.str.extract(r'(\(\d\d\d\d\))',expand=False)
#Removing the parentheses
movies_df['year'] = movies_df.year.str.extract(r'(\d\d\d\d)',expand=False)
#Removing the years from the 'title' column
movies_df['title'] = movies_df.title.str.replace(r'(\(\d\d\d\d\))', '', regex=True)
#Applying the strip function to get rid of any ending whitespace characters that may have appeared
movies_df['title'] = movies_df['title'].apply(lambda x: x.strip())

2.

old:

userSubsetGroup.get_group(1130)

new:

userSubsetGroup.get_group((1130,))

3.

این بخش به طوری کلی اجرا نمی شد، اما مشکل بسیار کوچک بود تنها با تبدیل tuple به int

old:

pearsonDF = pd.DataFrame.from_dict(pearsonCorrelationDict, orient='index')
pearsonDF.columns = ['similarityIndex']
pearsonDF['userId'] = pearsonDF.index
pearsonDF.index = range(len(pearsonDF))
pearsonDF.head()

new:

pearsonDF = pd.DataFrame.from_dict(pearsonCorrelationDict, orient='index')
pearsonDF.columns = ['similarityIndex']
pearsonDF['userId'] = pearsonDF.index
#Using int not tuple
pearsonDF["userId"] = pearsonDF["userId"].apply(lambda x: x[0])
pearsonDF.index = range(len(pearsonDF))
pearsonDF.head()

سپاس از شما بابت دوره بسیار عالی و کاربردی
خندون بمونید ... :)