数据分析作业四-基于用户及物品数据进行内容推荐

## 导入支持库
import pandas as pd
import matplotlib.pyplot as plt
import sklearn.metrics as metrics
import numpy as np
from sklearn.neighbors import NearestNeighbors
from scipy.spatial.distance import correlation
from sklearn.metrics.pairwise import pairwise_distances
import ipywidgets as widgets
from IPython.display import display, clear_output
from contextlib import contextmanager
import warnings
warnings.filterwarnings('ignore')
import numpy as np
import os, sys
import re
import seaborn as sns

## 加载数据集并检查书籍，用户和评级数据集的形状
books = pd.read_csv('F:\\data\\bleeding_data\\BX-Books.csv',
                    sep=None,encoding="latin-1")
books.columns = ['ISBN', 'bookTitle', 'bookAuthor',
                 'yearOfPublication', 'publisher',
                 'imageUrlS', 'imageUrlM', 'imageUrlL']

users = pd.read_csv('F:\\data\\bleeding_data\\BX-Users.csv',
                    sep=None, encoding="latin-1")
users.columns = ['userID', 'Location', 'Age']


ratings = pd.read_csv('F:\\data\\bleeding_data\\BX-Book-Ratings.csv',
                      sep=None, encoding="latin-1")
ratings.columns = ['userID', 'ISBN', 'bookRating']

print (books.shape)
print (users.shape)
print (ratings.shape)

(271360, 8)
(278858, 3)
(1149780, 3)

## 一、图书数据集
books.head()
ISBNbookTitlebookAuthoryearOfPublicationpublisherimageUrlSimageUrlMimageUrlL
00195153448Classical MythologyMark P. O. Morford2002Oxford University Presshttp://images.amazon.com/images/P/0195153448.0...http://images.amazon.com/images/P/0195153448.0...http://images.amazon.com/images/P/0195153448.0...
10002005018Clara CallanRichard Bruce Wright2001HarperFlamingo Canadahttp://images.amazon.com/images/P/0002005018.0...http://images.amazon.com/images/P/0002005018.0...http://images.amazon.com/images/P/0002005018.0...
20060973129Decision in NormandyCarlo D'Este1991HarperPerennialhttp://images.amazon.com/images/P/0060973129.0...http://images.amazon.com/images/P/0060973129.0...http://images.amazon.com/images/P/0060973129.0...
30374157065Flu: The Story of the Great Influenza Pandemic...Gina Bari Kolata1999Farrar Straus Girouxhttp://images.amazon.com/images/P/0374157065.0...http://images.amazon.com/images/P/0374157065.0...http://images.amazon.com/images/P/0374157065.0...
40393045218The Mummies of UrumchiE. J. W. Barber1999W. W. Norton & Companyhttp://images.amazon.com/images/P/0393045218.0...http://images.amazon.com/images/P/0393045218.0...http://images.amazon.com/images/P/0393045218.0...
## url不需要分析，进行删除
books.drop(['imageUrlS', 'imageUrlM', 'imageUrlL'],axis=1,inplace=True)
books.head()
ISBNbookTitlebookAuthoryearOfPublicationpublisher
00195153448Classical MythologyMark P. O. Morford2002Oxford University Press
10002005018Clara CallanRichard Bruce Wright2001HarperFlamingo Canada
20060973129Decision in NormandyCarlo D'Este1991HarperPerennial
30374157065Flu: The Story of the Great Influenza Pandemic...Gina Bari Kolata1999Farrar Straus Giroux
40393045218The Mummies of UrumchiE. J. W. Barber1999W. W. Norton & Company
## books.dtypes
books.dtypes

ISBN                 object
bookTitle            object
bookAuthor           object
yearOfPublication    object
publisher            object
dtype: object

## 现在检查属性的唯一值
books.bookTitle.unique()

array(['Classical Mythology', 'Clara Callan', 'Decision in Normandy', ...,
       'Lily Dale : The True Story of the Town that Talks to the Dead',
       "Republic (World's Classics)",
       "A Guided Tour of Rene Descartes' Meditations on First Philosophy with Complete Translations of the Meditations by Ronald Rubin"],
      dtype=object)

books.yearOfPublication.unique()

array(['2002', '2001', '1991', '1999', '2000', '1993', '1996', '1988',
       '2004', '1998', '1994', '2003', '1997', '1983', '1979', '1995',
       '1982', '1985', '1992', '1986', '1978', '1980', '1952', '1987',
       '1990', '1981', '1989', '1984', '0', '1968', '1961', '1958',
       '1974', '1976', '1971', '1977', '1975', '1965', '1941', '1970',
       '1962', '1973', '1972', '1960', '1966', '1920', '1956', '1959',
       '1953', '1951', '1942', '1963', '1964', '1969', '1954', '1950',
       '1967', '2005', '1957', '1940', '1937', '1955', '1946', '1936',
       '1930', '2011', '1925', '1948', '1943', '1947', '1945', '1923',
       '2020', '1939', '1926', '1938', '2030', '1911', '1904', '1949',
       '1932', '1928', '1929', '1927', '1931', '1914', '2050', '1934',
       '1910', '1933', '1902', '1924', '1921', '1900', '2038', '2026',
       '1944', '1917', '1901', '2010', '1908', '1906', '1935', '1806',
       '2021', '2012', '2006', 'DK Publishing Inc', 'Gallimard', '1909',
       '2008', '1378', '1919', '1922', '1897', '2024', '1376', '2037'],
      dtype=object)

books.loc[books.yearOfPublication == 'DK Publishing Inc',:]
books.yearOfPublication.unique()

array(['2002', '2001', '1991', '1999', '2000', '1993', '1996', '1988',
       '2004', '1998', '1994', '2003', '1997', '1983', '1979', '1995',
       '1982', '1985', '1992', '1986', '1978', '1980', '1952', '1987',
       '1990', '1981', '1989', '1984', '0', '1968', '1961', '1958',
       '1974', '1976', '1971', '1977', '1975', '1965', '1941', '1970',
       '1962', '1973', '1972', '1960', '1966', '1920', '1956', '1959',
       '1953', '1951', '1942', '1963', '1964', '1969', '1954', '1950',
       '1967', '2005', '1957', '1940', '1937', '1955', '1946', '1936',
       '1930', '2011', '1925', '1948', '1943', '1947', '1945', '1923',
       '2020', '1939', '1926', '1938', '2030', '1911', '1904', '1949',
       '1932', '1928', '1929', '1927', '1931', '1914', '2050', '1934',
       '1910', '1933', '1902', '1924', '1921', '1900', '2038', '2026',
       '1944', '1917', '1901', '2010', '1908', '1906', '1935', '1806',
       '2021', '2012', '2006', 'DK Publishing Inc', 'Gallimard', '1909',
       '2008', '1378', '1919', '1922', '1897', '2024', '1376', '2037'],
      dtype=object)

print(books.loc[books.yearOfPublication == 'DK Publishing Inc',:])

              ISBN                                          bookTitle  \
209538  078946697X  DK Readers: Creating the X-Men, How It All Beg...   
221678  0789466953  DK Readers: Creating the X-Men, How Comic Book...   

       bookAuthor  yearOfPublication  \
209538       2000  DK Publishing Inc   
221678       2000  DK Publishing Inc   

                                                publisher  
209538  http://images.amazon.com/images/P/078946697X.0...  
221678  http://images.amazon.com/images/P/0789466953.0...  

books.loc[books.yearOfPublication == 'DK Publishing Inc',:]
ISBNbookTitlebookAuthoryearOfPublicationpublisher
209538078946697XDK Readers: Creating the X-Men, How It All Beg...2000DK Publishing Inchttp://images.amazon.com/images/P/078946697X.0...
2216780789466953DK Readers: Creating the X-Men, How Comic Book...2000DK Publishing Inchttp://images.amazon.com/images/P/0789466953.0...
## 从上面可以看出，bookAuthor错误地装载了bookTitle，因此需要进行修正。
# ISBN '0789466953'
books.loc[books.ISBN == '0789466953','yearOfPublication'] = 2000
books.loc[books.ISBN == '0789466953','bookAuthor'] = "James Buckley"
books.loc[books.ISBN == '0789466953','publisher'] = "DK Publishing Inc"
books.loc[books.ISBN == '0789466953','bookTitle'] = "DK Readers: Creating the X-Men, How Comic Books Come to Life (Level 4: Proficient Readers)"

#ISBN '078946697X'
books.loc[books.ISBN == '078946697X','yearOfPublication'] = 2000
books.loc[books.ISBN == '078946697X','bookAuthor'] = "Michael Teitelbaum"
books.loc[books.ISBN == '078946697X','publisher'] = "DK Publishing Inc"
books.loc[books.ISBN == '078946697X','bookTitle'] = "DK Readers: Creating the X-Men, How It All Began (Level 4: Proficient Readers)"

books.loc[(books.ISBN == '0789466953') | (books.ISBN == '078946697X'),:]
ISBNbookTitlebookAuthoryearOfPublicationpublisher
209538078946697XDK Readers: Creating the X-Men, How It All Beg...Michael Teitelbaum2000DK Publishing Inc
2216780789466953DK Readers: Creating the X-Men, How Comic Book...James Buckley2000DK Publishing Inc
## 继续纠正出版年鉴的类型
books.yearOfPublication=pd.to_numeric(books.yearOfPublication, errors='coerce')
sorted(books['yearOfPublication'].unique())

[0.0,
 1376.0,
 1378.0,
 1806.0,
 1897.0,
 1900.0,
 1901.0,
 1902.0,
 1904.0,
 1906.0,
 1908.0,
 1909.0,
 1910.0,
 1911.0,
 1914.0,
 1917.0,
 1919.0,
 1920.0,
 1921.0,
 1922.0,
 1923.0,
 1924.0,
 1925.0,
 1926.0,
 1927.0,
 1928.0,
 1929.0,
 1930.0,
 1931.0,
 1932.0,
 1933.0,
 1934.0,
 1935.0,
 1936.0,
 1937.0,
 1938.0,
 1939.0,
 1940.0,
 1941.0,
 1942.0,
 1943.0,
 1944.0,
 1945.0,
 1946.0,
 1947.0,
 1948.0,
 1949.0,
 1950.0,
 1951.0,
 1952.0,
 1953.0,
 1954.0,
 1955.0,
 1956.0,
 1957.0,
 1958.0,
 1959.0,
 1960.0,
 1961.0,
 1962.0,
 1963.0,
 1964.0,
 1965.0,
 1966.0,
 1967.0,
 1968.0,
 1969.0,
 1970.0,
 1971.0,
 1972.0,
 1973.0,
 1974.0,
 1975.0,
 1976.0,
 1977.0,
 1978.0,
 1979.0,
 1980.0,
 1981.0,
 1982.0,
 1983.0,
 1984.0,
 1985.0,
 1986.0,
 1987.0,
 1988.0,
 1989.0,
 1990.0,
 1991.0,
 1992.0,
 1993.0,
 1994.0,
 1995.0,
 1996.0,
 1997.0,
 1998.0,
 1999.0,
 2000.0,
 2001.0,
 2002.0,
 2003.0,
 2004.0,
 2005.0,
 2006.0,
 2008.0,
 2010.0,
 2011.0,
 2012.0,
 2020.0,
 2021.0,
 2024.0,
 2026.0,
 2030.0,
 2037.0,
 2038.0,
 2050.0,
 nan]

## 现在可以看出yearOfPublication的类型为int，其值范围为0-2050。

## 由于该数据集建于2004年，我假设2006年之后的所有年份都无效，保留两年的保证金，以防数据集可能已更新。

## 对于所有无效条目（包括0），我将这些条目转换为NaN，然后​​用剩余年份的平均值替换它们。
books.loc[(books.yearOfPublication > 2006) | (books.yearOfPublication == 0),'yearOfPublication'] = np.NAN

# 用年出版的平均价值代替NaNs在案例数据集被更新的情况下保留一定的空白
books.yearOfPublication.fillna(round(books.yearOfPublication.mean()), inplace=True)
books.yearOfPublication.isnull().sum()

0

books.yearOfPublication = books.yearOfPublication.astype(np.int32)

## publisher
books.loc[books.publisher.isnull(),:]
ISBNbookTitlebookAuthoryearOfPublicationpublisher
128890193169656XTyrant MoonElaine Corvidae2002NaN
1290371931696993Finders KeepersLinnea Sinclair2001NaN
## 检查行是否有书签作为查找器，看看我们是否能得到任何线索

## 与不同的出版商和图书作者的所有行
books.loc[(books.bookTitle == 'Tyrant Moon'),:]
ISBNbookTitlebookAuthoryearOfPublicationpublisher
128890193169656XTyrant MoonElaine Corvidae2002NaN
books.loc[(books.bookTitle == 'Finders Keepers'),:]
ISBNbookTitlebookAuthoryearOfPublicationpublisher
10799082177364XFinders KeepersFern Michaels2002Zebra Books
420190070465037Finders KeepersBarbara Nickolae1989McGraw-Hill Companies
582640688118461Finders KeepersEmily Rodda1993Harpercollins Juvenile Books
666781575663236Finders KeepersFern Michaels1998Kensington Publishing Corporation
1290371931696993Finders KeepersLinnea Sinclair2001NaN
1343090156309505Finders KeepersWill1989Voyager Books
1734730973146907Finders KeepersSean M. Costello2002Red Tower Publications
1958850061083909Finders KeepersSharon Sala2003HarperTorch
2118740373261160Finders KeepersElizabeth Travis1993Worldwide Library
## 由图书作者检查以找到模式

## 都有不同的出版商。这里没有线索
books.loc[(books.bookAuthor == 'Elaine Corvidae'),:]
ISBNbookTitlebookAuthoryearOfPublicationpublisher
1267621931696934Winter's OrphansElaine Corvidae2001Novelbooks
128890193169656XTyrant MoonElaine Corvidae2002NaN
1290010759901880WolfkinElaine Corvidae2001Hard Shell Word Factory
## 由图书作者检查以找到模式
books.loc[(books.bookAuthor == 'Linnea Sinclair'),:]
ISBNbookTitlebookAuthoryearOfPublicationpublisher
1290371931696993Finders KeepersLinnea Sinclair2001NaN
## 因为没有什么共同的东西可以推断出NaNs的发布者，将它们替换为“other”
books.loc[(books.ISBN == '193169656X'),'publisher'] = 'other'
books.loc[(books.ISBN == '1931696993'),'publisher'] = 'other'

## 二、用户数据集
print (users.shape)
users.head()

(278858, 3)
userIDLocationAge
01nyc, new york, usaNaN
12stockton, california, usa18.0
23moscow, yukon territory, russiaNaN
34porto, v.n.gaia, portugal17.0
45farnborough, hants, united kingdomNaN
users.dtypes

userID        int64
Location     object
Age         float64
dtype: object

users.userID.values

array([     1,      2,      3, ..., 278856, 278857, 278858], dtype=int64)

## Age 
sorted(users.Age.unique())

[nan,
 0.0,
 1.0,
 2.0,
 3.0,
 4.0,
 5.0,
 6.0,
 7.0,
 8.0,
 9.0,
 10.0,
 11.0,
 12.0,
 13.0,
 14.0,
 15.0,
 16.0,
 17.0,
 18.0,
 19.0,
 20.0,
 21.0,
 22.0,
 23.0,
 24.0,
 25.0,
 26.0,
 27.0,
 28.0,
 29.0,
 30.0,
 31.0,
 32.0,
 33.0,
 34.0,
 35.0,
 36.0,
 37.0,
 38.0,
 39.0,
 40.0,
 41.0,
 42.0,
 43.0,
 44.0,
 45.0,
 46.0,
 47.0,
 48.0,
 49.0,
 50.0,
 51.0,
 52.0,
 53.0,
 54.0,
 55.0,
 56.0,
 57.0,
 58.0,
 59.0,
 60.0,
 61.0,
 62.0,
 63.0,
 64.0,
 65.0,
 66.0,
 67.0,
 68.0,
 69.0,
 70.0,
 71.0,
 72.0,
 73.0,
 74.0,
 75.0,
 76.0,
 77.0,
 78.0,
 79.0,
 80.0,
 81.0,
 82.0,
 83.0,
 84.0,
 85.0,
 86.0,
 87.0,
 88.0,
 89.0,
 90.0,
 91.0,
 92.0,
 93.0,
 94.0,
 95.0,
 96.0,
 97.0,
 98.0,
 99.0,
 100.0,
 101.0,
 102.0,
 103.0,
 104.0,
 105.0,
 106.0,
 107.0,
 108.0,
 109.0,
 110.0,
 111.0,
 113.0,
 114.0,
 115.0,
 116.0,
 118.0,
 119.0,
 123.0,
 124.0,
 127.0,
 128.0,
 132.0,
 133.0,
 136.0,
 137.0,
 138.0,
 140.0,
 141.0,
 143.0,
 146.0,
 147.0,
 148.0,
 151.0,
 152.0,
 156.0,
 157.0,
 159.0,
 162.0,
 168.0,
 172.0,
 175.0,
 183.0,
 186.0,
 189.0,
 199.0,
 200.0,
 201.0,
 204.0,
 207.0,
 208.0,
 209.0,
 210.0,
 212.0,
 219.0,
 220.0,
 223.0,
 226.0,
 228.0,
 229.0,
 230.0,
 231.0,
 237.0,
 239.0,
 244.0]

## 年龄栏有一些无效的条目，比如nan，0和非常高的值，比如100和以上
users.loc[(users.Age > 90) | (users.Age < 5), 'Age'] = np.nan

## 用平均值代替NaN
## 将数据类型设置为int
users.Age = users.Age.fillna(users.Age.mean())
users.Age = users.Age.astype(np.int32)

sorted(users.Age.unique())

[5,
 6,
 7,
 8,
 9,
 10,
 11,
 12,
 13,
 14,
 15,
 16,
 17,
 18,
 19,
 20,
 21,
 22,
 23,
 24,
 25,
 26,
 27,
 28,
 29,
 30,
 31,
 32,
 33,
 34,
 35,
 36,
 37,
 38,
 39,
 40,
 41,
 42,
 43,
 44,
 45,
 46,
 47,
 48,
 49,
 50,
 51,
 52,
 53,
 54,
 55,
 56,
 57,
 58,
 59,
 60,
 61,
 62,
 63,
 64,
 65,
 66,
 67,
 68,
 69,
 70,
 71,
 72,
 73,
 74,
 75,
 76,
 77,
 78,
 79,
 80,
 81,
 82,
 83,
 84,
 85,
 86,
 87,
 88,
 89,
 90]

## 三、评级数据集
ratings.shape

(1149780, 3)

## 如果每个用户对每个条目进行评级，那么评级数据集将有nusers * nbooks条目，这表明数据集非常稀疏。
n_users = users.shape[0]
n_books = books.shape[0]
print (n_users * n_books)

75670906880

ratings.head(5)
userIDISBNbookRating
0276725034545104X0
127672601550612245
227672704465208020
3276729052165615X3
427672905217950286
ratings.bookRating.unique()

array([ 0,  5,  3,  6,  8,  7, 10,  9,  4,  1,  2], dtype=int64)

ratings_new = ratings[ratings.ISBN.isin(books.ISBN)]

print (ratings.shape)
print (ratings_new.shape)

(1149780, 3)
(1031136, 3)

## 没有新用户添加，因此我们将使用高于数据集的新用户（1031136，3）
print ("number of users: " + str(n_users))
print ("number of books: " + str(n_books))

number of users: 278858
number of books: 271360

sparsity=1.0-len(ratings_new)/float(n_users*n_books)
print ('图书交叉数据集的稀疏级别是 ' +  str(sparsity*100) + ' %')

图书交叉数据集的稀疏级别是 99.99863734155898 %

ratings.bookRating.unique()

array([ 0,  5,  3,  6,  8,  7, 10,  9,  4,  1,  2], dtype=int64)

ratings_explicit = ratings_new[ratings_new.bookRating != 0]
ratings_implicit = ratings_new[ratings_new.bookRating == 0]

print (ratings_new.shape)
print( ratings_explicit.shape)
print (ratings_implicit.shape)

(1031136, 3)
(383842, 3)
(647294, 3)

## 统计
sns.countplot(data=ratings_explicit , x='bookRating')
plt.show()

在这里插入图片描述

## 基于简单流行度的推荐系统
ratings_count = pd.DataFrame(ratings_explicit.groupby(['ISBN'])['bookRating'].sum())
top10 = ratings_count.sort_values('bookRating', ascending = False).head(10)
print ("推荐下列书籍")
top10.merge(books, left_index = True, right_on = 'ISBN')

推荐下列书籍
bookRatingISBNbookTitlebookAuthoryearOfPublicationpublisher
40857870316666343The Lovely Bones: A NovelAlice Sebold2002Little, Brown
74841080385504209The Da Vinci CodeDan Brown2003Doubleday
52231340312195516The Red Tent (Bestselling Backlist)Anita Diamant1998Picador USA
21432798059035342XHarry Potter and the Sorcerer's Stone (Harry P...J. K. Rowling1999Arthur A. Levine Books
35625950142001740The Secret Life of BeesSue Monk Kidd2003Penguin Books
2625510971880107Wild AnimusRich Shapero2004Too Far
110525240060928336Divine Secrets of the Ya-Ya Sisterhood: A NovelRebecca Wells1997Perennial
70624020446672211Where the Heart Is (Oprah's Book Club (Paperba...Billie Letts1998Warner Books
23122190452282152Girl with a Pearl EarringTracy Chevalier2001Plume Books
11821790671027360Angels &amp; DemonsDan Brown2001Pocket Star
users_exp_ratings = users[users.userID.isin(ratings_explicit.userID)]
users_imp_ratings = users[users.userID.isin(ratings_implicit.userID)]

print (users.shape)
print (users_exp_ratings.shape)
print (users_imp_ratings.shape)

(278858, 3)
(68091, 3)
(52451, 3)

## 基于协同过滤的推荐系统
counts1 = ratings_explicit['userID'].value_counts()
ratings_explicit = ratings_explicit[ratings_explicit['userID'].isin(counts1[counts1 >= 100].index)]
counts = ratings_explicit['bookRating'].value_counts()
ratings_explicit = ratings_explicit[ratings_explicit['bookRating'].isin(counts[counts >= 100].index)]

ratings_matrix = ratings_explicit.pivot(index='userID', columns='ISBN', values='bookRating')
userID = ratings_matrix.index
ISBN = ratings_matrix.columns
print(ratings_matrix.shape)
ratings_matrix.head()

(449, 66574)
ISBN00009131540001046438000104687X00010472130001047973000104799X0001048082000105373600010537440001055607...B000092Q0AB00009EF82B00009NDANB0000DYXIDB0000T6KHIB0000VZEJQB0000X8HIEB00013AX9EB0001I1KOGB000234N3A
userID
2033NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
2110NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
2276NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
4017NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
4385NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN

5 rows × 66574 columns

n_users = ratings_matrix.shape[0] #只考虑那些给出明确评级的用户
n_books = ratings_matrix.shape[1]
print (n_users, n_books)

449 66574

ratings_matrix.fillna(0, inplace = True)
ratings_matrix = ratings_matrix.astype(np.int32)

ratings_matrix.head(5)
ISBN00009131540001046438000104687X00010472130001047973000104799X0001048082000105373600010537440001055607...B000092Q0AB00009EF82B00009NDANB0000DYXIDB0000T6KHIB0000VZEJQB0000X8HIEB00013AX9EB0001I1KOGB000234N3A
userID
20330000000000...0000000000
21100000000000...0000000000
22760000000000...0000000000
40170000000000...0000000000
43850000000000...0000000000

5 rows × 66574 columns

sparsity=1.0-len(ratings_explicit)/float(users_exp_ratings.shape[0]*n_books)
print ('图书交叉数据集的稀疏级别是 ' +  str(sparsity*100) + ' %')

图书交叉数据集的稀疏级别是 99.99772184106935 %

## 基于用户的协同过滤
global metric,k
k=10
metric='cosine'
def findksimilarusers(user_id, ratings, metric = metric, k=k):
    similarities=[]
    indices=[]
    model_knn = NearestNeighbors(metric = metric, algorithm = 'brute') 
    model_knn.fit(ratings)
    loc = ratings.index.get_loc(user_id)
    distances, indices = model_knn.kneighbors(ratings.iloc[loc, :].values.reshape(1, -1), n_neighbors = k+1)
    similarities = 1-distances.flatten()
            
    return similarities,indices

def predict_userbased(user_id, item_id, ratings, metric = metric, k=k):
    prediction=0
    user_loc = ratings.index.get_loc(user_id)
    item_loc = ratings.columns.get_loc(item_id)
    similarities, indices=findksimilarusers(user_id, ratings,metric, k) #similar users based on cosine similarity
    mean_rating = ratings.iloc[user_loc,:].mean() #to adjust for zero based indexing
    sum_wt = np.sum(similarities)-1
    product=1
    wtd_sum = 0 
    
    for i in range(0, len(indices.flatten())):
        if indices.flatten()[i] == user_loc:
            continue;
        else: 
            ratings_diff = ratings.iloc[indices.flatten()[i],item_loc]-np.mean(ratings.iloc[indices.flatten()[i],:])
            product = ratings_diff * (similarities[i])
            wtd_sum = wtd_sum + product
    
    #在非常稀疏的数据集的情况下，使用基于协作的方法的相关度量可能会给出负面的评价
    #在这里的处理如下
    if prediction <= 0:
        prediction = 1   
    elif prediction >10:
        prediction = 10
    
    prediction = int(round(mean_rating + (wtd_sum/sum_wt)))
    print ('用户预测等级 {0} -> item {1}: {2}'.format(user_id,item_id,prediction))
 
    return prediction

## 测试
predict_userbased(11676,'0001056107',ratings_matrix)

用户预测等级 11676 -> item 0001056107: 2





2

## 基于项目的协同过滤
def findksimilaritems(item_id, ratings, metric=metric, k=k):
    similarities=[]
    indices=[]
    ratings=ratings.T
    loc = ratings.index.get_loc(item_id)
    model_knn = NearestNeighbors(metric = metric, algorithm = 'brute')
    model_knn.fit(ratings)
    
    distances, indices = model_knn.kneighbors(ratings.iloc[loc, :].values.reshape(1, -1), n_neighbors = k+1)
    similarities = 1-distances.flatten()
 
    return similarities,indices

def predict_itembased(user_id, item_id, ratings, metric = metric, k=k):
    prediction= wtd_sum =0
    user_loc = ratings.index.get_loc(user_id)
    item_loc = ratings.columns.get_loc(item_id)
    similarities, indices=findksimilaritems(item_id, ratings) #similar users based on correlation coefficients
    sum_wt = np.sum(similarities)-1
    product=1
    for i in range(0, len(indices.flatten())):
        if indices.flatten()[i] == item_loc:
            continue;
        else:
            product = ratings.iloc[user_loc,indices.flatten()[i]] * (similarities[i])
            wtd_sum = wtd_sum + product                              
    prediction = int(round(wtd_sum/sum_wt))
    
    #在非常稀疏的数据集的情况下，使用基于协作的方法的相关度量可能会给出负面的评价
    #在这里处理的是下面的//代码，没有下面的代码片段，下面的代码片段是为了避免负面影响
    #在使用相关度规时，可能会出现非常稀疏的数据集的预测
    if prediction <= 0:
        prediction = 1   
    elif prediction >10:
        prediction = 10
 
    print ('用户预测等级 {0} -> item {1}: {2}'.format(user_id,item_id,prediction)    )  
    
    return prediction

## 测试
prediction = predict_itembased(11676,'0001056107',ratings_matrix)

用户预测等级 11676 -> item 0001056107: 1

本文来自互联网用户投稿，该文观点仅代表作者本人，不代表本站立场。本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如若转载，请注明出处：/a/94357.html

如若内容造成侵权/违法违规/事实不符，请联系我们进行投诉反馈qq邮箱809451989@qq.com，一经查实，立即删除！

相关文章

死信队列理解与使用

死信队列理解与使用

一、简介 在rabbitMQ中常用的交换机有三种&#xff0c;直连交换机、广播交换机、主题交换机&#xff1b; 直连交换机中队列与交换机需要约定好routingKey去进行绑定&#xff1b; 广播交换机并不需要routingKey绑定,只需队列与交换机绑定即可&#xff1b; 主题交换机最大的特…
阅读更多...
monorepo更新组件报错，提示“无法加载文件 C:\Program Files\nodejs\pnpm.ps1，因为在此系统上禁止运行脚本”

monorepo更新组件报错，提示“无法加载文件 C:\Program Files\nodejs\pnpm.ps1，因为在此系统上禁止运行脚本”

解决方法&#xff1a; 第一步&#xff1a;管理员身份运行 window.powershell&#xff0c; win x打开powerShell命令框&#xff0c;进入到对应项目路径。 第二步&#xff1a;执行&#xff1a;get-ExecutionPolicy&#xff0c;显示Restricted&#xff0c;表示状态是禁止的; 第…
阅读更多...
TCP协议的重点知识点

TCP协议的重点知识点

TCP协议的重点知识点 TCP(传输控制协议)是一种面向连接、可靠的数据传输协议,工作在传输层,提供可靠的字节流服务。它是互联网协议栈中最重要、最复杂的协议之一,也是面试中常被问到的知识点。本文将详细介绍TCP协议的各个重要概念。 TCP基本特性 TCP主要具有以下基本特性: …
阅读更多...
如何用 QGIS 下载高清天地图影像机，同时解决下载质量差的问题！

如何用 QGIS 下载高清天地图影像机，同时解决下载质量差的问题！

使用 QGIS 我们可以获得下面这种图像，既有大范围，又有更高的细节（地图级别），基本上把整个苏州市中心城区的建筑物都囊括进去了。 还可以下载大范围、高清晰度的各种在线卫星底图服务的影像，比如大面积的哨兵2影像，但国外的服务器一般都很烂，不可能是电信、移动的问题，…
阅读更多...
python爬虫12：实战4

python爬虫12：实战4

python爬虫12&#xff1a;实战4 前言 ​ python实现网络爬虫非常简单&#xff0c;只需要掌握一定的基础知识和一定的库使用技巧即可。本系列目标旨在梳理相关知识点&#xff0c;方便以后复习。 申明 ​ 本系列所涉及的代码仅用于个人研究与讨论&#xff0c;并不会对网站产生不好…
阅读更多...
QGIS 如何添加天地图

QGIS 如何添加天地图

相信很多小伙伴在 QGIS 里面添加天地图的时候一定感觉很困惑，按照官网的操作申请 Key 之后，添加相对应的服务地址之后看不到地图或者地图不正常显示，今天我们就来解决这个问题 以下所有操作基于 QGIS 3.22 版本 申请 Key 1. 添加天地图的第一步需要申请 Key，首先要注册天…
阅读更多...
Git基础教程-常用命令整理：学会Git使用方法和错误解决

Git基础教程-常用命令整理：学会Git使用方法和错误解决

目录 一、了解Git的基本概念 二、Git的安装和配置 Git的安装 Git的配置 用户信息 文本编辑器 差异分析工具 查看配置信息 三、Git的基本操作 基本原理 基本操作命令 基本操作示例 场景一&#xff1a;创建新仓库 场景二&#xff1a;拉取并编辑远程仓库 四、常见问…
阅读更多...
MySQL之事务与引擎

MySQL之事务与引擎

目录 一、事物 1、事务的概念 2、事务的ACID特点 3、事务之间的相互影响 4、Mysql及事务隔离级别(四种) 5、演示 1、查询会话事务隔离级别 2、查询会话事务隔离级别 3、设置全局事务隔离级别 4、设置会话事务隔离级别 6、事务控制语句 7、演示 1、测试提交事务 2、测试事…
阅读更多...
countDown+react+hook

countDown+react+hook

道阻且长&#xff0c;行而不辍&#xff0c;未来可期 知识点一&#xff1a; new Date().getTime()可以得到得到1970年01月1日0点零分以来的毫秒数。单位是毫秒 new Date().getTime()/1000获取秒数1分钟60秒&#xff0c;1小时60分钟1hour:60*60>单位是秒 60*60*1000>单位…
阅读更多...
Java的成员类可以被private修饰

Java的成员类可以被private修饰

说明 Java的成员类可以被private修饰&#xff0c;但外部类、局部类不能被private修饰。 示例 成员类用private修饰—允许 下面代码中的成员类Class2 被private修饰&#xff0c;是允许的&#xff1a; package com.thb;public class Parent {public class Class1 { }private…
阅读更多...
ChatGPT Prompting开发实战（一）

ChatGPT Prompting开发实战（一）

一、关于ChatGPT Prompting概述 当我们使用ChatGPT或者调用OpenAI的API时&#xff0c;就是在使用prompt进行交互&#xff0c;用户在对话过程中输入的一切信息都是prompt&#xff08;提示词&#xff09;&#xff0c;当然工业级的prompt与人们通常理解的prompt可能不太一样。下面…
阅读更多...
基于java swing和mysql实现的仓库商品管理系统（源码+数据库+运行指导视频）

基于java swing和mysql实现的仓库商品管理系统（源码+数据库+运行指导视频）

一、项目简介 本项目是一套基于java swing和mysql实现的仓库商品管理系统&#xff0c;主要针对计算机相关专业的正在做毕设的学生与需要项目实战练习的Java学习者。 包含&#xff1a;项目源码、项目文档、数据库脚本等&#xff0c;该项目附带全部源码可作为毕设使用。 项目都经…
阅读更多...
新SDK平台下载开源全志V853的SDK

新SDK平台下载开源全志V853的SDK

获取SDK SDK 使用 Repo 工具管理&#xff0c;拉取 SDK 需要配置安装 Repo 工具。 Repo is a tool built on top of Git. Repo helps manage many Git repositories, does the uploads to revision control systems, and automates parts of the development workflow. Repo is…
阅读更多...
若依vue打印的简单方法

若依vue打印的简单方法

像我们后端程序员做前端的话,有时候真不需要知道什么原理,直接塞就好了 我们选用基于hiprint 的vue-plugin-hiprint来打印 目的是为了实现点击某些行的数据,然后点击某个按钮直接弹出下面的打印 此链接 大佬是原创,我拿来总结梳理一下 插件进阶功能请移步: 链接 插件模板制作页…
阅读更多...
Leetcode每日一题：1267. 统计参与通信的服务器（2023.8.24 C++）

Leetcode每日一题：1267. 统计参与通信的服务器（2023.8.24 C++）

目录 1267. 统计参与通信的服务器 题目描述&#xff1a; 实现代码与解析&#xff1a; 写法一&#xff1a;两次遍历 hash 原理思路&#xff1a; 写法二&#xff1a;三次遍历 原理思路&#xff1a; 1267. 统计参与通信的服务器 题目描述&#xff1a; 这里有一幅服务器分…
阅读更多...
三维模型OBJ格式轻量化压缩并行计算处理方法浅析

三维模型OBJ格式轻量化压缩并行计算处理方法浅析

三维模型OBJ格式轻量化压缩并行计算处理方法浅析 三维模型的轻量化是指通过一系列技术和算法来减小三维模型的文件大小&#xff0c;以提高模型在计算机中的加载、渲染和传输效率。并行计算是利用多个计算单元同时执行任务&#xff0c;以加速计算过程的一种技术。在三维模型的O…
阅读更多...
基于Spring Boot的软件缺陷追踪系统的设计与实现（Java+spring boot+MySQL）

基于Spring Boot的软件缺陷追踪系统的设计与实现（Java+spring boot+MySQL）

获取源码或者论文请私信博主 演示视频&#xff1a; 基于Spring Boot的软件缺陷追踪系统的设计与实现&#xff08;Javaspring bootMySQL&#xff09; 使用技术&#xff1a; 前端&#xff1a;html css javascript jQuery ajax thymeleaf 微信小程序 后端&#xff1a;Java spri…
阅读更多...
【数据分析】统计量

【数据分析】统计量

1. 均值、众数描述数据的集中趋势度量&#xff0c;四分位差、极差描述数据的离散程度。 2. 标准差、四分位差、异众比率度量离散程度&#xff0c;协方差是度量相关性。 期望值分别为E[X]与E[Y]的两个实随机变量X与Y之间的协方差Cov(X,Y)定义为&#xff1a; 从直观上来看&…
阅读更多...
ppt转pdf免费的工具哪个好用？ppt在线转pdf的方法分享

ppt转pdf免费的工具哪个好用？ppt在线转pdf的方法分享

在工作和学习中&#xff0c;将PPT文件转换为PDF格式具有重要意义。PDF文件的大小较小&#xff0c;适用于各种平台和设备&#xff0c;保持了原始文件的内容和格式&#xff0c;具有广泛的可读性和兼容性。那么小编就来为大家详细地说一说“ppt转pdf免费的工具哪个好用?ppt在线转…
阅读更多...
US-DAS1、US-P2A单路及双路插头式比例放大器

US-DAS1、US-P2A单路及双路插头式比例放大器

US-P1、US-P2A、US-P2F插头式安装比例放大器控制不带电反馈的双路比例电磁铁的比例阀&#xff0c;如直动式或先导式比例方向阀的驱动控制。 工作电源24VDC标准&#xff1b; 兼容指令10V、4-20mA、0~10V、0~5V(电位器控制&#xff09;&#xff1b; 输出电流0~2A&#xff1b; …
阅读更多...
最新文章

Copyright @ 2022~2023