Chopped B
Shunya
2017年12月11日
## Make numerical values into categorical in order to complement market basket analysis
rest[["weight"]] <- ordered(cut(rest[["weight"]], c(0,66,120)), labels = c("Small", "Large"))
rest[["height"]] <- ordered(cut(rest[["height"]], c(0,1.65,2)), labels = c("short", "tall"))
rest[["birth_year"]] <- ordered(cut(rest[["birth_year"]], c(1930,1986,1989,1994)),
labels = c("old", "middle","young"))
rest <- rest[-18:-20]
rest$rating <- recode(rest$rating, "'0' = 'bad';'1'='normal';'2'='good'")
rest$rating = as.factor(rest$rating)
rest$food_rating = as.factor(rest$food_rating)
rest$service_rating = as.factor(rest$service_rating)
rest1 <- as(rest, "transactions")
summary(rest1)## transactions as itemMatrix in sparse format with
## 4090 rows (elements/itemsets/transactions) and
## 313 columns (items) and a density of 0.06727232
##
## most frequent items:
## marital_status=single activity=student
## 3919 3655
## religion=Catholic Upayment=cash
## 3606 3352
## dress_preference=informal (Other)
## 2651 68937
##
## element (itemset/transaction) length distribution:
## sizes
## 20 21 22
## 68 3724 298
##
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 20.00 21.00 21.00 21.06 21.00 22.00
##
## includes extended item information - examples:
## labels variables levels
## 1 userID=U1001 userID U1001
## 2 userID=U1002 userID U1002
## 3 userID=U1003 userID U1003
##
## includes extended transaction information - examples:
## transactionID
## 1 1
## 2 2
## 3 3aa=as(rest1,"matrix") # transforms transaction matrix into incidence matrix
aa[1:2,] # print the first two rows of the incidence matrix## userID=U1001 userID=U1002 userID=U1003 userID=U1004 userID=U1005
## 1 TRUE FALSE FALSE FALSE FALSE
## 2 TRUE FALSE FALSE FALSE FALSE
## userID=U1006 userID=U1007 userID=U1008 userID=U1009 userID=U1010
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1011 userID=U1012 userID=U1013 userID=U1014 userID=U1015
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1016 userID=U1017 userID=U1018 userID=U1019 userID=U1020
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1021 userID=U1022 userID=U1023 userID=U1026 userID=U1027
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1028 userID=U1029 userID=U1030 userID=U1031 userID=U1032
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1033 userID=U1034 userID=U1035 userID=U1036 userID=U1037
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1038 userID=U1039 userID=U1040 userID=U1041 userID=U1042
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1043 userID=U1044 userID=U1045 userID=U1046 userID=U1047
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1048 userID=U1049 userID=U1050 userID=U1051 userID=U1052
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1053 userID=U1054 userID=U1055 userID=U1056 userID=U1057
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1058 userID=U1059 userID=U1060 userID=U1061 userID=U1062
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1063 userID=U1064 userID=U1065 userID=U1066 userID=U1067
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1068 userID=U1069 userID=U1070 userID=U1071 userID=U1072
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1073 userID=U1074 userID=U1075 userID=U1076 userID=U1077
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1078 userID=U1079 userID=U1080 userID=U1081 userID=U1082
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1083 userID=U1084 userID=U1085 userID=U1086 userID=U1087
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1089 userID=U1090 userID=U1091 userID=U1092 userID=U1093
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1094 userID=U1095 userID=U1096 userID=U1097 userID=U1098
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1099 userID=U1100 userID=U1101 userID=U1102 userID=U1103
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1104 userID=U1105 userID=U1106 userID=U1107 userID=U1108
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1109 userID=U1110 userID=U1111 userID=U1112 userID=U1113
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1114 userID=U1115 userID=U1116 userID=U1117 userID=U1118
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1119 userID=U1120 userID=U1121 userID=U1123 userID=U1124
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1125 userID=U1126 userID=U1127 userID=U1128 userID=U1129
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1131 userID=U1132 userID=U1133 userID=U1134 userID=U1135
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1136 userID=U1137 userID=U1138 smoker drink_level=abstemious
## 1 FALSE FALSE FALSE FALSE TRUE
## 2 FALSE FALSE FALSE FALSE TRUE
## drink_level=casual drinker drink_level=social drinker dress_preference=?
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## dress_preference=elegant dress_preference=formal
## 1 FALSE FALSE
## 2 FALSE FALSE
## dress_preference=informal dress_preference=no preference ambience=?
## 1 TRUE FALSE FALSE
## 2 TRUE FALSE FALSE
## ambience=family ambience=friends ambience=solitary transport=?
## 1 TRUE FALSE FALSE FALSE
## 2 TRUE FALSE FALSE FALSE
## transport=car owner transport=on foot transport=public marital_status=?
## 1 FALSE TRUE FALSE FALSE
## 2 FALSE TRUE FALSE FALSE
## marital_status=married marital_status=single marital_status=widow
## 1 FALSE TRUE FALSE
## 2 FALSE TRUE FALSE
## hijos=? hijos=dependent hijos=independent hijos=kids birth_year=old
## 1 FALSE FALSE TRUE FALSE FALSE
## 2 FALSE FALSE TRUE FALSE FALSE
## birth_year=middle birth_year=young interest=eco-friendly interest=none
## 1 TRUE FALSE FALSE FALSE
## 2 TRUE FALSE FALSE FALSE
## interest=retro interest=technology interest=variety
## 1 FALSE FALSE TRUE
## 2 FALSE FALSE TRUE
## personality=conformist personality=hard-worker
## 1 FALSE FALSE
## 2 FALSE FALSE
## personality=hunter-ostentatious personality=thrifty-protector
## 1 FALSE TRUE
## 2 FALSE TRUE
## religion=Catholic religion=Christian religion=Jewish religion=Mormon
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## religion=none activity=? activity=professional activity=student
## 1 TRUE FALSE FALSE TRUE
## 2 TRUE FALSE FALSE TRUE
## activity=unemployed activity=working-class color=black color=blue
## 1 FALSE FALSE TRUE FALSE
## 2 FALSE FALSE TRUE FALSE
## color=green color=orange color=purple color=red color=white color=yellow
## 1 FALSE FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE FALSE
## weight=Small weight=Large budget=? budget=high budget=low budget=medium
## 1 FALSE TRUE FALSE FALSE FALSE TRUE
## 2 FALSE TRUE FALSE FALSE FALSE TRUE
## height=short height=tall rating=bad rating=good rating=normal
## 1 FALSE TRUE FALSE FALSE TRUE
## 2 FALSE TRUE FALSE FALSE TRUE
## food_rating=0 food_rating=1 food_rating=2 service_rating=0
## 1 FALSE TRUE FALSE FALSE
## 2 FALSE TRUE FALSE FALSE
## service_rating=1 service_rating=2 Rcuisine=Afghan Rcuisine=African
## 1 TRUE FALSE FALSE FALSE
## 2 TRUE FALSE FALSE FALSE
## Rcuisine=American Rcuisine=Armenian Rcuisine=Asian Rcuisine=Australian
## 1 TRUE FALSE FALSE FALSE
## 2 TRUE FALSE FALSE FALSE
## Rcuisine=Austrian Rcuisine=Bagels Rcuisine=Bakery Rcuisine=Bar
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Bar_Pub_Brewery Rcuisine=Barbecue Rcuisine=Basque
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Brazilian Rcuisine=Breakfast-Brunch Rcuisine=British
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Burgers Rcuisine=Burmese Rcuisine=Cafe-Coffee_Shop
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Cafeteria Rcuisine=Cajun-Creole Rcuisine=California
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Cambodian Rcuisine=Canadian Rcuisine=Caribbean Rcuisine=Chilean
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Chinese Rcuisine=Contemporary Rcuisine=Continental-European
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Cuban Rcuisine=Deli-Sandwiches Rcuisine=Dessert-Ice_Cream
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Dim_Sum Rcuisine=Diner Rcuisine=Doughnuts
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Dutch-Belgian Rcuisine=Eastern_European Rcuisine=Eclectic
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Ethiopian Rcuisine=Family Rcuisine=Fast_Food Rcuisine=Filipino
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Fine_Dining Rcuisine=French Rcuisine=Fusion Rcuisine=Game
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=German Rcuisine=Greek Rcuisine=Hawaiian Rcuisine=Hot_Dogs
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Hungarian Rcuisine=Indian-Pakistani Rcuisine=Indigenous
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Indonesian Rcuisine=International Rcuisine=Irish
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Israeli Rcuisine=Italian Rcuisine=Jamaican Rcuisine=Japanese
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Juice Rcuisine=Korean Rcuisine=Kosher Rcuisine=Latin_American
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Lebanese Rcuisine=Malaysian Rcuisine=Mediterranean
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Mexican Rcuisine=Middle_Eastern Rcuisine=Mongolian
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Moroccan Rcuisine=North_African Rcuisine=Organic-Healthy
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Pacific_Northwest Rcuisine=Pacific_Rim Rcuisine=Persian
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Peruvian Rcuisine=Pizzeria Rcuisine=Polish Rcuisine=Polynesian
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Portuguese Rcuisine=Regional Rcuisine=Romanian
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Russian-Ukrainian Rcuisine=Scandinavian Rcuisine=Seafood
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Soup Rcuisine=Southeast_Asian Rcuisine=Southern
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Southwestern Rcuisine=Spanish Rcuisine=Steaks Rcuisine=Sushi
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Swiss Rcuisine=Tapas Rcuisine=Tea_House Rcuisine=Tex-Mex
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Thai Rcuisine=Tibetan Rcuisine=Tunisian Rcuisine=Turkish
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Vegetarian Rcuisine=Vietnamese Upayment=American_Express
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Upayment=bank_debit_cards Upayment=cash Upayment=MasterCard-Eurocard
## 1 FALSE TRUE FALSE
## 2 FALSE TRUE FALSE
## Upayment=VISA
## 1 FALSE
## 2 FALSEitemFrequencyPlot(rest1[, itemFrequency(rest1) > 0.4], cex.names = 1)
rules <- apriori(rest1, parameter = list(support = 0.05, confidence = 0.6))## Apriori
##
## Parameter specification:
## confidence minval smax arem aval originalSupport maxtime support minlen
## 0.6 0.1 1 none FALSE TRUE 5 0.05 1
## maxlen target ext
## 10 rules FALSE
##
## Algorithmic control:
## filter tree heap memopt load sort verbose
## 0.1 TRUE TRUE FALSE TRUE 2 TRUE
##
## Absolute minimum support count: 204
##
## set item appearances ...[0 item(s)] done [0.00s].
## set transactions ...[313 item(s), 4090 transaction(s)] done [0.00s].
## sorting and recoding items ... [55 item(s)] done [0.00s].
## creating transaction tree ... done [0.00s].
## checking subsets of size 1 2 3 4 5 6 7 8 9 10 done [0.57s].
## writing ... [8543938 rule(s)] done [1.81s].
## creating S4 object ... done [5.43s].summary(rules)## set of 8543938 rules
##
## rule length distribution (lhs + rhs):sizes
## 1 2 3 4 5 6 7 8 9
## 6 640 8986 50545 177902 470514 995145 1708171 2391765
## 10
## 2740264
##
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 1.000 8.000 9.000 8.603 10.000 10.000
##
## summary of quality measures:
## support confidence lift count
## Min. :0.05012 Min. :0.6000 Min. : 0.6714 Min. : 205
## 1st Qu.:0.07922 1st Qu.:1.0000 1st Qu.: 1.5847 1st Qu.: 324
## Median :0.35257 Median :1.0000 Median : 2.0328 Median :1442
## Mean :0.24601 Mean :0.9817 Mean : 2.2350 Mean :1006
## 3rd Qu.:0.35257 3rd Qu.:1.0000 3rd Qu.: 2.5138 3rd Qu.:1442
## Max. :0.95819 Max. :1.0000 Max. :10.2262 Max. :3919
##
## mining info:
## data ntransactions support confidence
## rest1 4090 0.05 0.6showing most frequent customer information that appered with bad restaurant rating.
rulesrating0 <- subset(rules, subset = rhs %in% "rating=bad" & lift > 1.2)
inspect(sort(rulesrating0, by = "support")[1:10])## lhs rhs support confidence lift count
## [1] {service_rating=0} => {rating=bad} 0.4283619 0.8897918 2.012859 1752
## [2] {food_rating=0} => {rating=bad} 0.4261614 0.9467681 2.141749 1743
## [3] {food_rating=0,
## service_rating=0} => {rating=bad} 0.4217604 0.9862779 2.231126 1725
## [4] {marital_status=single,
## service_rating=0} => {rating=bad} 0.4207824 0.8921721 2.018243 1721
## [5] {marital_status=single,
## food_rating=0} => {rating=bad} 0.4193154 0.9464680 2.141070 1715
## [6] {service_rating=0,
## Upayment=cash} => {rating=bad} 0.4185819 0.9164882 2.073250 1712
## [7] {food_rating=0,
## Upayment=cash} => {rating=bad} 0.4166259 0.9514238 2.152281 1704
## [8] {activity=student,
## service_rating=0} => {rating=bad} 0.4156479 0.8975713 2.030457 1700
## [9] {marital_status=single,
## food_rating=0,
## service_rating=0} => {rating=bad} 0.4149144 0.9866279 2.231918 1697
## [10] {marital_status=single,
## activity=student,
## service_rating=0} => {rating=bad} 0.4139364 0.8976670 2.030674 1693showing most frequent customer information that appered with good restaurant rating.
rulesrationg2 <- subset(rules, subset = rhs %in% "rating=good" & lift > 1.2)
inspect(sort(rulesrationg2, by = "support")[1:10])## lhs rhs support confidence lift count
## [1] {food_rating=2} => {rating=good} 0.2447433 0.7170487 2.331263 1001
## [2] {marital_status=single,
## food_rating=2} => {rating=good} 0.2242054 0.7158470 2.327356 917
## [3] {hijos=independent,
## food_rating=2} => {rating=good} 0.2202934 0.7208000 2.343459 901
## [4] {service_rating=2} => {rating=good} 0.2166259 0.8635478 2.807560 886
## [5] {marital_status=single,
## hijos=independent,
## food_rating=2} => {rating=good} 0.2156479 0.7194127 2.338949 882
## [6] {budget=medium,
## food_rating=2} => {rating=good} 0.2019560 0.7108434 2.311089 826
## [7] {marital_status=single,
## service_rating=2} => {rating=good} 0.1990220 0.8724544 2.836517 814
## [8] {hijos=independent,
## service_rating=2} => {rating=good} 0.1948655 0.8865406 2.882314 797
## [9] {marital_status=single,
## hijos=independent,
## service_rating=2} => {rating=good} 0.1916870 0.8868778 2.883410 784
## [10] {religion=Catholic,
## food_rating=2} => {rating=good} 0.1909535 0.6948399 2.259058 781showing most frequent customer information that appered with bad food rating.
foodrating0 <- subset(rules, subset = rhs %in% "food_rating=0" & lift > 1.2)
inspect(sort(foodrating0, by = "support")[1:10])## lhs rhs support confidence lift count
## [1] {service_rating=0} => {food_rating=0} 0.4276284 0.8882682 1.973393 1749
## [2] {rating=bad} => {food_rating=0} 0.4261614 0.9640487 2.141749 1743
## [3] {rating=bad,
## service_rating=0} => {food_rating=0} 0.4217604 0.9845890 2.187381 1725
## [4] {marital_status=single,
## service_rating=0} => {food_rating=0} 0.4205379 0.8916537 1.980915 1720
## [5] {marital_status=single,
## rating=bad} => {food_rating=0} 0.4193154 0.9716714 2.158683 1715
## [6] {weight=Small} => {food_rating=0} 0.4188264 0.6636962 1.474480 1713
## [7] {service_rating=0,
## Upayment=cash} => {food_rating=0} 0.4178484 0.9148822 2.032519 1709
## [8] {rating=bad,
## Upayment=cash} => {food_rating=0} 0.4166259 0.9726027 2.160752 1704
## [9] {marital_status=single,
## weight=Small} => {food_rating=0} 0.4158924 0.6845070 1.520714 1701
## [10] {marital_status=single,
## rating=bad,
## service_rating=0} => {food_rating=0} 0.4149144 0.9860546 2.190637 1697showing most frequent customer information that appered withgood food rating.
foodrating2 <- subset(rules, subset = rhs %in% "food_rating=2" & lift > 1.2)
inspect(sort(foodrating2, by = "support")[1:10])## lhs rhs support confidence lift count
## [1] {hijos=independent,
## budget=medium} => {food_rating=2} 0.2528117 0.6032672 1.767452 1034
## [2] {marital_status=single,
## hijos=independent,
## budget=medium} => {food_rating=2} 0.2469438 0.6026253 1.765571 1010
## [3] {rating=good} => {food_rating=2} 0.2447433 0.7957075 2.331263 1001
## [4] {marital_status=single,
## rating=good} => {food_rating=2} 0.2242054 0.7851027 2.300194 917
## [5] {hijos=independent,
## rating=good} => {food_rating=2} 0.2202934 0.8066249 2.363249 901
## [6] {marital_status=single,
## hijos=independent,
## rating=good} => {food_rating=2} 0.2156479 0.8047445 2.357740 882
## [7] {service_rating=2} => {food_rating=2} 0.2039120 0.8128655 2.381533 834
## [8] {religion=Catholic,
## budget=medium,
## height=tall} => {food_rating=2} 0.2029340 0.6120944 1.793314 830
## [9] {budget=medium,
## rating=good} => {food_rating=2} 0.2019560 0.8227092 2.410373 826
## [10] {personality=thrifty-protector,
## activity=student} => {food_rating=2} 0.1997555 0.6069837 1.778340 817showing most frequent customer information that appered with bad service rating.
servicerating0 <- subset(rules, subset = rhs %in% "service_rating=0" & lift > 1.2)
inspect(sort(servicerating0, by = "support")[1:10])## lhs rhs support confidence lift count
## [1] {marital_status=single,
## religion=Catholic,
## Upayment=cash} => {service_rating=0} 0.4339853 0.6045640 1.255798 1775
## [2] {religion=Catholic,
## activity=student,
## Upayment=cash} => {service_rating=0} 0.4288509 0.6083940 1.263754 1754
## [3] {marital_status=single,
## religion=Catholic,
## activity=student,
## Upayment=cash} => {service_rating=0} 0.4286064 0.6146564 1.276762 1753
## [4] {rating=bad} => {service_rating=0} 0.4283619 0.9690265 2.012859 1752
## [5] {food_rating=0} => {service_rating=0} 0.4276284 0.9500272 1.973393 1749
## [6] {rating=bad,
## food_rating=0} => {service_rating=0} 0.4217604 0.9896730 2.055745 1725
## [7] {marital_status=single,
## rating=bad} => {service_rating=0} 0.4207824 0.9750708 2.025414 1721
## [8] {marital_status=single,
## food_rating=0} => {service_rating=0} 0.4205379 0.9492274 1.971732 1720
## [9] {weight=Small} => {service_rating=0} 0.4198044 0.6652460 1.381847 1717
## [10] {rating=bad,
## Upayment=cash} => {service_rating=0} 0.4185819 0.9771689 2.029772 1712showing most frequent customer information that appered with good service rating.
servicerating2 <- subset(rules, subset = rhs %in% "service_rating=2" & lift > 1.2)
inspect(sort(servicerating2, by = "support")[1:10])## lhs rhs support confidence lift count
## [1] {rating=good} => {service_rating=2} 0.2166259 0.7042925 2.807560 886
## [2] {marital_status=single,
## rating=good} => {service_rating=2} 0.1990220 0.6969178 2.778162 814
## [3] {hijos=independent,
## rating=good} => {service_rating=2} 0.1948655 0.7135184 2.844337 797
## [4] {marital_status=single,
## hijos=independent,
## rating=good} => {service_rating=2} 0.1916870 0.7153285 2.851553 784
## [5] {rating=good,
## food_rating=2} => {service_rating=2} 0.1841076 0.7522478 2.998726 753
## [6] {budget=medium,
## rating=good} => {service_rating=2} 0.1841076 0.7500000 2.989766 753
## [7] {budget=medium,
## food_rating=2} => {service_rating=2} 0.1706601 0.6006885 2.394557 698
## [8] {religion=Catholic,
## rating=good} => {service_rating=2} 0.1689487 0.6861966 2.735423 691
## [9] {marital_status=single,
## budget=medium,
## rating=good} => {service_rating=2} 0.1674817 0.7453754 2.971331 685
## [10] {marital_status=single,
## rating=good,
## food_rating=2} => {service_rating=2} 0.1667482 0.7437296 2.964770 682Visualization
##Bad restaurant rating
plotly_arules(rulesrating0)##Good restaurant rating
plotly_arules(rulesrationg2)Serch table for the result of market basket analysis(restaurant rating)
inspectDT(rulesrationg2)The customers who rated as good in food and service, they are likely to rate restaurant rating as good.
And if the individual is independent and single, and has midium budget, restaurant rating is tend to be good.
Without food and servise rating
rest <- rest[-19:-20]
rest1 <- as(rest, "transactions")
rest1## transactions in sparse format with
## 4090 transactions (rows) and
## 307 items (columns)summary(rest1)## transactions as itemMatrix in sparse format with
## 4090 rows (elements/itemsets/transactions) and
## 307 columns (items) and a density of 0.06207243
##
## most frequent items:
## marital_status=single activity=student
## 3919 3655
## religion=Catholic Upayment=cash
## 3606 3352
## dress_preference=informal (Other)
## 2651 60757
##
## element (itemset/transaction) length distribution:
## sizes
## 18 19 20
## 68 3724 298
##
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 18.00 19.00 19.00 19.06 19.00 20.00
##
## includes extended item information - examples:
## labels variables levels
## 1 userID=U1001 userID U1001
## 2 userID=U1002 userID U1002
## 3 userID=U1003 userID U1003
##
## includes extended transaction information - examples:
## transactionID
## 1 1
## 2 2
## 3 3aa=as(rest1,"matrix") # transforms transaction matrix into incidence matrix
aa[1:2,] # print the first two rows of the incidence matrix## userID=U1001 userID=U1002 userID=U1003 userID=U1004 userID=U1005
## 1 TRUE FALSE FALSE FALSE FALSE
## 2 TRUE FALSE FALSE FALSE FALSE
## userID=U1006 userID=U1007 userID=U1008 userID=U1009 userID=U1010
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1011 userID=U1012 userID=U1013 userID=U1014 userID=U1015
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1016 userID=U1017 userID=U1018 userID=U1019 userID=U1020
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1021 userID=U1022 userID=U1023 userID=U1026 userID=U1027
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1028 userID=U1029 userID=U1030 userID=U1031 userID=U1032
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1033 userID=U1034 userID=U1035 userID=U1036 userID=U1037
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1038 userID=U1039 userID=U1040 userID=U1041 userID=U1042
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1043 userID=U1044 userID=U1045 userID=U1046 userID=U1047
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1048 userID=U1049 userID=U1050 userID=U1051 userID=U1052
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1053 userID=U1054 userID=U1055 userID=U1056 userID=U1057
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1058 userID=U1059 userID=U1060 userID=U1061 userID=U1062
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1063 userID=U1064 userID=U1065 userID=U1066 userID=U1067
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1068 userID=U1069 userID=U1070 userID=U1071 userID=U1072
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1073 userID=U1074 userID=U1075 userID=U1076 userID=U1077
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1078 userID=U1079 userID=U1080 userID=U1081 userID=U1082
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1083 userID=U1084 userID=U1085 userID=U1086 userID=U1087
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1089 userID=U1090 userID=U1091 userID=U1092 userID=U1093
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1094 userID=U1095 userID=U1096 userID=U1097 userID=U1098
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1099 userID=U1100 userID=U1101 userID=U1102 userID=U1103
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1104 userID=U1105 userID=U1106 userID=U1107 userID=U1108
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1109 userID=U1110 userID=U1111 userID=U1112 userID=U1113
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1114 userID=U1115 userID=U1116 userID=U1117 userID=U1118
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1119 userID=U1120 userID=U1121 userID=U1123 userID=U1124
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1125 userID=U1126 userID=U1127 userID=U1128 userID=U1129
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1131 userID=U1132 userID=U1133 userID=U1134 userID=U1135
## 1 FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE
## userID=U1136 userID=U1137 userID=U1138 smoker drink_level=abstemious
## 1 FALSE FALSE FALSE FALSE TRUE
## 2 FALSE FALSE FALSE FALSE TRUE
## drink_level=casual drinker drink_level=social drinker dress_preference=?
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## dress_preference=elegant dress_preference=formal
## 1 FALSE FALSE
## 2 FALSE FALSE
## dress_preference=informal dress_preference=no preference ambience=?
## 1 TRUE FALSE FALSE
## 2 TRUE FALSE FALSE
## ambience=family ambience=friends ambience=solitary transport=?
## 1 TRUE FALSE FALSE FALSE
## 2 TRUE FALSE FALSE FALSE
## transport=car owner transport=on foot transport=public marital_status=?
## 1 FALSE TRUE FALSE FALSE
## 2 FALSE TRUE FALSE FALSE
## marital_status=married marital_status=single marital_status=widow
## 1 FALSE TRUE FALSE
## 2 FALSE TRUE FALSE
## hijos=? hijos=dependent hijos=independent hijos=kids birth_year=old
## 1 FALSE FALSE TRUE FALSE FALSE
## 2 FALSE FALSE TRUE FALSE FALSE
## birth_year=middle birth_year=young interest=eco-friendly interest=none
## 1 TRUE FALSE FALSE FALSE
## 2 TRUE FALSE FALSE FALSE
## interest=retro interest=technology interest=variety
## 1 FALSE FALSE TRUE
## 2 FALSE FALSE TRUE
## personality=conformist personality=hard-worker
## 1 FALSE FALSE
## 2 FALSE FALSE
## personality=hunter-ostentatious personality=thrifty-protector
## 1 FALSE TRUE
## 2 FALSE TRUE
## religion=Catholic religion=Christian religion=Jewish religion=Mormon
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## religion=none activity=? activity=professional activity=student
## 1 TRUE FALSE FALSE TRUE
## 2 TRUE FALSE FALSE TRUE
## activity=unemployed activity=working-class color=black color=blue
## 1 FALSE FALSE TRUE FALSE
## 2 FALSE FALSE TRUE FALSE
## color=green color=orange color=purple color=red color=white color=yellow
## 1 FALSE FALSE FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE FALSE FALSE
## weight=Small weight=Large budget=? budget=high budget=low budget=medium
## 1 FALSE TRUE FALSE FALSE FALSE TRUE
## 2 FALSE TRUE FALSE FALSE FALSE TRUE
## height=short height=tall rating=bad rating=good rating=normal
## 1 FALSE TRUE FALSE FALSE TRUE
## 2 FALSE TRUE FALSE FALSE TRUE
## Rcuisine=Afghan Rcuisine=African Rcuisine=American Rcuisine=Armenian
## 1 FALSE FALSE TRUE FALSE
## 2 FALSE FALSE TRUE FALSE
## Rcuisine=Asian Rcuisine=Australian Rcuisine=Austrian Rcuisine=Bagels
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Bakery Rcuisine=Bar Rcuisine=Bar_Pub_Brewery Rcuisine=Barbecue
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Basque Rcuisine=Brazilian Rcuisine=Breakfast-Brunch
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=British Rcuisine=Burgers Rcuisine=Burmese
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Cafe-Coffee_Shop Rcuisine=Cafeteria Rcuisine=Cajun-Creole
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=California Rcuisine=Cambodian Rcuisine=Canadian
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Caribbean Rcuisine=Chilean Rcuisine=Chinese
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Contemporary Rcuisine=Continental-European Rcuisine=Cuban
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Deli-Sandwiches Rcuisine=Dessert-Ice_Cream Rcuisine=Dim_Sum
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Diner Rcuisine=Doughnuts Rcuisine=Dutch-Belgian
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Eastern_European Rcuisine=Eclectic Rcuisine=Ethiopian
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Family Rcuisine=Fast_Food Rcuisine=Filipino
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Fine_Dining Rcuisine=French Rcuisine=Fusion Rcuisine=Game
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=German Rcuisine=Greek Rcuisine=Hawaiian Rcuisine=Hot_Dogs
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Hungarian Rcuisine=Indian-Pakistani Rcuisine=Indigenous
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Indonesian Rcuisine=International Rcuisine=Irish
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Israeli Rcuisine=Italian Rcuisine=Jamaican Rcuisine=Japanese
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Juice Rcuisine=Korean Rcuisine=Kosher Rcuisine=Latin_American
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Lebanese Rcuisine=Malaysian Rcuisine=Mediterranean
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Mexican Rcuisine=Middle_Eastern Rcuisine=Mongolian
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Moroccan Rcuisine=North_African Rcuisine=Organic-Healthy
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Pacific_Northwest Rcuisine=Pacific_Rim Rcuisine=Persian
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Peruvian Rcuisine=Pizzeria Rcuisine=Polish Rcuisine=Polynesian
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Portuguese Rcuisine=Regional Rcuisine=Romanian
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Russian-Ukrainian Rcuisine=Scandinavian Rcuisine=Seafood
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Soup Rcuisine=Southeast_Asian Rcuisine=Southern
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Rcuisine=Southwestern Rcuisine=Spanish Rcuisine=Steaks Rcuisine=Sushi
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Swiss Rcuisine=Tapas Rcuisine=Tea_House Rcuisine=Tex-Mex
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Thai Rcuisine=Tibetan Rcuisine=Tunisian Rcuisine=Turkish
## 1 FALSE FALSE FALSE FALSE
## 2 FALSE FALSE FALSE FALSE
## Rcuisine=Vegetarian Rcuisine=Vietnamese Upayment=American_Express
## 1 FALSE FALSE FALSE
## 2 FALSE FALSE FALSE
## Upayment=bank_debit_cards Upayment=cash Upayment=MasterCard-Eurocard
## 1 FALSE TRUE FALSE
## 2 FALSE TRUE FALSE
## Upayment=VISA
## 1 FALSE
## 2 FALSEitemFrequencyPlot(rest1[, itemFrequency(rest1) > 0.4], cex.names = 1)
rules <- apriori(rest1, parameter = list(support = 0.05, confidence = 0.6))## Apriori
##
## Parameter specification:
## confidence minval smax arem aval originalSupport maxtime support minlen
## 0.6 0.1 1 none FALSE TRUE 5 0.05 1
## maxlen target ext
## 10 rules FALSE
##
## Algorithmic control:
## filter tree heap memopt load sort verbose
## 0.1 TRUE TRUE FALSE TRUE 2 TRUE
##
## Absolute minimum support count: 204
##
## set item appearances ...[0 item(s)] done [0.00s].
## set transactions ...[307 item(s), 4090 transaction(s)] done [0.00s].
## sorting and recoding items ... [49 item(s)] done [0.00s].
## creating transaction tree ... done [0.00s].
## checking subsets of size 1 2 3 4 5 6 7 8 9 10 done [0.17s].
## writing ... [2927841 rule(s)] done [0.67s].
## creating S4 object ... done [1.60s].rules## set of 2927841 rulessummary(rules)## set of 2927841 rules
##
## rule length distribution (lhs + rhs):sizes
## 1 2 3 4 5 6 7 8 9 10
## 6 528 6451 32093 99211 229506 424879 635081 763754 736332
##
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 1.000 7.000 9.000 8.304 10.000 10.000
##
## summary of quality measures:
## support confidence lift count
## Min. :0.05012 Min. :0.6000 Min. : 0.6714 Min. : 205.0
## 1st Qu.:0.07922 1st Qu.:1.0000 1st Qu.: 1.5428 1st Qu.: 324.0
## Median :0.35257 Median :1.0000 Median : 2.0328 Median :1442.0
## Mean :0.23523 Mean :0.9846 Mean : 2.2743 Mean : 962.1
## 3rd Qu.:0.35257 3rd Qu.:1.0000 3rd Qu.: 2.6576 3rd Qu.:1442.0
## Max. :0.95819 Max. :1.0000 Max. :10.2262 Max. :3919.0
##
## mining info:
## data ntransactions support confidence
## rest1 4090 0.05 0.6rulesrating0 <- subset(rules, subset = rhs %in% "rating=bad" & lift > 1.2)
inspect(sort(rulesrating0, by = "support")[1:10])## lhs rhs support confidence lift count
## [1] {weight=Small} => {rating=bad} 0.4129584 0.6543975 1.480357 1689
## [2] {marital_status=single,
## weight=Small} => {rating=bad} 0.4070905 0.6700201 1.515698 1665
## [3] {weight=Small,
## Upayment=cash} => {rating=bad} 0.4051345 0.7066098 1.598470 1657
## [4] {activity=student,
## weight=Small} => {rating=bad} 0.4044010 0.7241681 1.638190 1654
## [5] {drink_level=casual drinker} => {rating=bad} 0.4026895 0.7005530 1.584769 1647
## [6] {drink_level=casual drinker,
## marital_status=single} => {rating=bad} 0.4026895 0.7005530 1.584769 1647
## [7] {religion=Catholic,
## weight=Small} => {rating=bad} 0.4026895 0.7245930 1.639151 1647
## [8] {marital_status=single,
## weight=Small,
## Upayment=cash} => {rating=bad} 0.4024450 0.7175240 1.623160 1646
## [9] {activity=student,
## weight=Small,
## Upayment=cash} => {rating=bad} 0.4007335 0.7463570 1.688385 1639
## [10] {marital_status=single,
## activity=student,
## weight=Small} => {rating=bad} 0.4002445 0.7430776 1.680967 1637rulesrationg2 <- subset(rules, subset = rhs %in% "rating=good" & lift > 1.2)
inspect(sort(rulesrationg2, by = "support")[1:10])## lhs rhs support confidence lift count
## [1] {birth_year=young,
## religion=Catholic,
## budget=medium} => {rating=good} 0.09975550 0.6089552 1.979831 408
## [2] {marital_status=single,
## birth_year=young,
## religion=Catholic,
## budget=medium} => {rating=good} 0.09975550 0.6144578 1.997721 408
## [3] {birth_year=young,
## activity=student,
## budget=medium} => {rating=good} 0.09804401 0.6039157 1.963446 401
## [4] {marital_status=single,
## birth_year=young,
## activity=student,
## budget=medium} => {rating=good} 0.09731051 0.6048632 1.966527 398
## [5] {hijos=independent,
## birth_year=young,
## budget=medium} => {rating=good} 0.09462103 0.6231884 2.026105 387
## [6] {marital_status=single,
## hijos=independent,
## birth_year=young,
## budget=medium} => {rating=good} 0.09462103 0.6231884 2.026105 387
## [7] {birth_year=young,
## height=tall} => {rating=good} 0.09388753 0.6104928 1.984830 384
## [8] {birth_year=young,
## budget=medium,
## Upayment=cash} => {rating=good} 0.09364303 0.6031496 1.960955 383
## [9] {birth_year=young,
## religion=Catholic,
## activity=student,
## budget=medium} => {rating=good} 0.09364303 0.6197411 2.014898 383
## [10] {marital_status=single,
## birth_year=young,
## religion=Catholic,
## activity=student,
## budget=medium} => {rating=good} 0.09364303 0.6197411 2.014898 383Without food and service rating, the result shows that if the customer is young, Cayholic, has the midim budget, student, and single, that customer is tend to rate the restaurant as good.
Visualization
plotly_arules(rulesrating0)plotly_arules(rulesrationg2)plot(rulesrationg2,main="food ratings' confidence vs. support")
inspectDT(rulesrationg2)