R ile Birliktelik Kuralları Analizi (Association Rules Analysis with R Project)
Merhaba,
Bir önceki yazımda Birliktelik Kuralları Analizi (Association Rules Analysis) ilgili R kodlarını paylaşacağımı belirtmiştim. Dilerseniz hızlıca başlayalım. R üzerinde Birliktelik Kuralları Analizi için kullanacağım kütüphaneler başlıca kütüphaneler: arules ve arulesViz
arules kütüphanesi ile Apriori algoritmasını kullanmak için, arulesViz kütüphanesi ile de Apriori algoritması sonucunda ortaya çıkan kuralları(rules) görselleştirmek için kullanacağım.
# Association Rules Analysis with R Project ----------------------------
# Clean the System & Console Variable --------------------
rm(list = ls())
cat("\014")
options(warn = -1)
# Sys_Date <- format(Sys.Date(), "%Y%m%d")
# Sys_Time <- format(Sys.time(), "%H:%M:%S")
# Install & Use Library ---------------------------------------------------
if (require('arules') == FALSE){
install.packages('arules')
library(arules)
}
if (require('arulesViz') == FALSE){
install.packages('arulesViz')
library(arulesViz)
}
if (require('htmlwidgets') == FALSE){
install.packages('htmlwidgets')
library(htmlwidgets)
}
# Set the System Path & Variable
if (require('here') == FALSE){
install.packages('here')
library(here)
}
Path <- here()
setwd(Path)arules ve arulesViz kütüphanelerinin dışında çalışma dizini için here ve interaktif çıktıları export edebilmemiz için htmlwigdets kütüphaneleri kullandım.
Şimdi kullanacağım veri seti içeriye alalım. Kullanacağımız veri seti aşağıdaki bağlantılardan indirebilirsiniz.
- http://www.sci.csueastbay.edu/~esuess/classes/Statistics_6620/Presentations/ml13/groceries.csv
- R içerisinde arules kütüphanesi ile birlikte gelen Groceries veri setini data(Groceries) yazılarak içeriye aktarılabilirsiniz.
# Load Dataset
Groceries = read.transactions(file = "./Input/groceries.csv",
format = c("basket"),
sep= ",",
cols = NULL,
rm.duplicates = TRUE,
encoding = "UTF-8")
# Second Way Data Load
# data(Groceries)Veri seti için bazı tanımlayıcı bilgilere bakacak olursak;
# Show Type and Class of Dataset typeof(Groceries) [1] "S4" class(Groceries) [1] "transactions" attr(,"package") [1] "arules" # Dimension of Dataset dim(Groceries) [1] 9835 169 length(Groceries) [1] 9835 str(Groceries) Formal class 'transactions' [package "arules"] with 3 slots ..@ data :Formal class 'ngCMatrix' [package "Matrix"] with 5 slots .. .. ..@ i : int [1:43367] 29 88 118 132 33 157 167 166 38 91 ... .. .. ..@ p : int [1:9836] 0 4 7 8 12 16 21 22 27 28 ... .. .. ..@ Dim : int [1:2] 169 9835 .. .. ..@ Dimnames:List of 2 .. .. .. ..$ : NULL .. .. .. ..$ : NULL .. .. ..@ factors : list() ..@ itemInfo :'data.frame': 169 obs. of 1 variable: .. ..$ labels: chr [1:169] "abrasive cleaner" "artif. sweetener" "baby cosmetics" "baby food" ... ..@ itemsetInfo:'data.frame': 0 obs. of 0 variables
Veri seti 9385×169 yani 9385 transaction(element) ve 169 items ‘tan oluşmaktadır. Veri seti içerisindeki ilk 10 gözleme bakalım;
# Show The Dataset
# inspect(Groceries)
inspect(Groceries[1:10])
items
[1] {citrus fruit,margarine,ready soups,semi-finished bread}
[2] {coffee,tropical fruit,yogurt}
[3] {whole milk}
[4] {cream cheese,meat spreads,pip fruit,yogurt}
[5] {condensed milk,long life bakery product,other vegetables,whole milk}
[6] {abrasive cleaner,butter,rice,whole milk,yogurt}
[7] {rolls/buns}
[8] {bottled beer,liquor (appetizer),other vegetables,rolls/buns,UHT-milk}
[9] {potted plants}
[10] {cereals,whole milk}Veri seti içerisindeki yer alan tüm items listeleyelim.
# Show the Items
ItemSetList <- Groceries@itemInfo
ItemSetList
labels
1 abrasive cleaner
2 artif. sweetener
3 baby cosmetics
4 baby food
5 bags
...
...
...
165 white bread
166 white wine
167 whole milk
168 yogurt
169 zwiebackİlk ve son 10 items listeyecek olursak;
# First 5 items of DataSet
# head(ItemSetList, n=10)
head(Groceries@itemInfo, n=10)
labels
1 abrasive cleaner
2 artif. sweetener
3 baby cosmetics
4 baby food
5 bags
6 baking powder
7 bathroom cleaner
8 beef
9 berries
10 beverages
# End 5 items of DataSet
# tail(ItemSetList, n=10)
tail(Groceries@itemInfo, n=10)
labels
160 UHT-milk
161 vinegar
162 waffles
163 whipped/sour cream
164 whisky
165 white bread
166 white wine
167 whole milk
168 yogurt
169 zwiebackVeri setini özetleyici bilgiler ile devam ediyoruz;
# Summary of Dataset
summary(Groceries)
transactions as itemMatrix in sparse format with
9835 rows (elements/itemsets/transactions) and
169 columns (items) and a density of 0.02609146
most frequent items:
whole milk other vegetables rolls/buns soda yogurt (Other)
2513 1903 1809 1715 1372 34055
element (itemset/transaction) length distribution:
sizes
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
2159 1643 1299 1005 855 645 545 438 350 246 182 117 78 77 55 46 29 14 14 9 11 4
23 24 26 27 28 29 32
6 1 1 1 1 3 1
Min. 1st Qu. Median Mean 3rd Qu. Max.
1.000 2.000 3.000 4.409 6.000 32.000
includes extended item information - examples:
labels
1 abrasive cleaner
2 artif. sweetener
3 baby cosmetics
En çok kullanılan items‘lar;
# Most Frequent Items
s@itemSummary
whole milk other vegetables rolls/buns soda yogurt (Other)
2513 1903 1809 1715 1372 34055En çok kullanılan 5 items‘a baktığımızda 2513 adet ile Whole Milk ilk sırada olduğu ve onu 1903 adet ile other vegetables izlediği görülmektedir.
Items (Birlikte yer alanda 1 ile 32 arasındaki items’ların) uzunlukları;
# Length Distribution Size s@lengths sizes 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 2159 1643 1299 1005 855 645 545 438 350 246 182 117 78 77 55 46 29 14 14 9 11 4 23 24 26 27 28 29 32 6 1 1 1 1 3 1
Bura tabloyu inceleyecek olursak; sadece 1 items‘ı içeren 2159 adet transaction olduğu ve sadece 32 items‘ı içeren 1 adet transaction olduğunu görülmektedir.
itemFrequency(Groceries, type = "relative")
abrasive cleaner artif. sweetener baby cosmetics baby food
0.0035587189 0.0032536858 0.0006100661 0.0001016777
bags baking powder bathroom cleaner beef
0.0004067107 0.0176919166 0.0027452974 0.0524656838
berries beverages bottled beer bottled water
0.0332486019 0.0260294865 0.0805287239 0.1105236401
brandy brown bread butter butter milk
0.0041687850 0.0648703610 0.0554143366 0.0279613625
cake bar candles candy canned beer
0.0132180986 0.0089476360 0.0298932384 0.0776817489
canned fish canned fruit canned vegetables cat food
0.0150482969 0.0032536858 0.0107778343 0.0232841891
cereals chewing gum chicken chocolate
0.0056939502 0.0210472801 0.0429079817 0.0496187087
chocolate marshmallow citrus fruit cleaner cling film/bags
0.0090493137 0.0827656329 0.0050838841 0.0113879004
cocoa drinks coffee condensed milk cooking chocolate
0.0022369090 0.0580579563 0.0102694459 0.0025419420
cookware cream cream cheese curd
0.0027452974 0.0013218099 0.0396542959 0.0532791052
curd cheese decalcifier dental care dessert
0.0050838841 0.0015251652 0.0057956279 0.0371123538
detergent dish cleaner dishes dog food
0.0192170819 0.0104728012 0.0175902389 0.0085409253
domestic eggs female sanitary products finished products fish
0.0634468734 0.0061006609 0.0065073716 0.0029486528
flour flower (seeds) flower soil/fertilizer frankfurter
0.0173868836 0.0103711235 0.0019318760 0.0589730554
frozen chicken frozen dessert frozen fish frozen fruits
0.0006100661 0.0107778343 0.0116929334 0.0012201322
frozen meals frozen potato products frozen vegetables fruit/vegetable juice
0.0283680732 0.0084392476 0.0480935435 0.0722928317
grapes hair spray ham hamburger meat
0.0223690900 0.0011184545 0.0260294865 0.0332486019
hard cheese herbs honey house keeping products
0.0245043213 0.0162684291 0.0015251652 0.0083375699
hygiene articles ice cream instant coffee Instant food products
0.0329435689 0.0250127097 0.0074224708 0.0080325369
jam ketchup kitchen towels kitchen utensil
0.0053889171 0.0042704626 0.0059989832 0.0004067107
light bulbs liqueur liquor liquor (appetizer)
0.0041687850 0.0009150991 0.0110828673 0.0079308592
liver loaf long life bakery product make up remover male cosmetics
0.0050838841 0.0374173869 0.0008134215 0.0045754957
margarine mayonnaise meat meat spreads
0.0585663447 0.0091509914 0.0258261312 0.0042704626
misc. beverages mustard napkins newspapers
0.0283680732 0.0119979664 0.0523640061 0.0798169802
nut snack nuts/prunes oil onions
0.0031520081 0.0033553635 0.0280630402 0.0310116929
organic products organic sausage other vegetables packaged fruit/vegetables
0.0016268429 0.0022369090 0.1934926284 0.0130147433
pasta pastry pet care photo/film
0.0150482969 0.0889679715 0.0094560244 0.0092526690
pickled vegetables pip fruit popcorn pork
0.0178952720 0.0756481952 0.0072191154 0.0576512456
potato products potted plants preservation products processed cheese
0.0028469751 0.0172852059 0.0002033554 0.0165734621
prosecco pudding powder ready soups red/blush wine
0.0020335536 0.0023385867 0.0018301983 0.0192170819
rice roll products rolls/buns root vegetables
0.0076258261 0.0102694459 0.1839349263 0.1089984748
rubbing alcohol rum salad dressing salt
0.0010167768 0.0044738180 0.0008134215 0.0107778343
salty snack sauces sausage seasonal products
0.0378240976 0.0054905948 0.0939501779 0.0142348754
semi-finished bread shopping bags skin care sliced cheese
0.0176919166 0.0985256736 0.0035587189 0.0245043213
snack products soap soda soft cheese
0.0030503305 0.0026436197 0.1743772242 0.0170818505
softener sound storage medium soups sparkling wine
0.0054905948 0.0001016777 0.0068124047 0.0055922725
specialty bar specialty cheese specialty chocolate specialty fat
0.0273512964 0.0085409253 0.0304016268 0.0036603965
specialty vegetables spices spread cheese sugar
0.0017285206 0.0051855618 0.0111845450 0.0338586680
sweet spreads syrup tea tidbits
0.0090493137 0.0032536858 0.0038637519 0.0023385867
toilet cleaner tropical fruit turkey UHT-milk
0.0007117438 0.1049313676 0.0081342145 0.0334519573
vinegar waffles whipped/sour cream whisky
0.0065073716 0.0384341637 0.0716827656 0.0008134215
white bread white wine whole milk yogurt
0.0420945602 0.0190137265 0.2555160142 0.1395017794
zwieback
0.0069140824Her bir items için relative değelerini gözükmektedir.
itemFrequency(Groceries, type = "absolute")
abrasive cleaner artif. sweetener baby cosmetics baby food
35 32 6 1
bags baking powder bathroom cleaner beef
4 174 27 516
berries beverages bottled beer bottled water
327 256 792 1087
brandy brown bread butter butter milk
41 638 545 275
cake bar candles candy canned beer
130 88 294 764
canned fish canned fruit canned vegetables cat food
148 32 106 229
cereals chewing gum chicken chocolate
56 207 422 488
chocolate marshmallow citrus fruit cleaner cling film/bags
89 814 50 112
cocoa drinks coffee condensed milk cooking chocolate
22 571 101 25
cookware cream cream cheese curd
27 13 390 524
curd cheese decalcifier dental care dessert
50 15 57 365
detergent dish cleaner dishes dog food
189 103 173 84
domestic eggs female sanitary products finished products fish
624 60 64 29
flour flower (seeds) flower soil/fertilizer frankfurter
171 102 19 580
frozen chicken frozen dessert frozen fish frozen fruits
6 106 115 12
frozen meals frozen potato products frozen vegetables fruit/vegetable juice
279 83 473 711
grapes hair spray ham hamburger meat
220 11 256 327
hard cheese herbs honey house keeping products
241 160 15 82
hygiene articles ice cream instant coffee Instant food products
324 246 73 79
jam ketchup kitchen towels kitchen utensil
53 42 59 4
light bulbs liqueur liquor liquor (appetizer)
41 9 109 78
liver loaf long life bakery product make up remover male cosmetics
50 368 8 45
margarine mayonnaise meat meat spreads
576 90 254 42
misc. beverages mustard napkins newspapers
279 118 515 785
nut snack nuts/prunes oil onions
31 33 276 305
organic products organic sausage other vegetables packaged fruit/vegetables
16 22 1903 128
pasta pastry pet care photo/film
148 875 93 91
pickled vegetables pip fruit popcorn pork
176 744 71 567
potato products potted plants preservation products processed cheese
28 170 2 163
prosecco pudding powder ready soups red/blush wine
20 23 18 189
rice roll products rolls/buns root vegetables
75 101 1809 1072
rubbing alcohol rum salad dressing salt
10 44 8 106
salty snack sauces sausage seasonal products
372 54 924 140
semi-finished bread shopping bags skin care sliced cheese
174 969 35 241
snack products soap soda soft cheese
30 26 1715 168
softener sound storage medium soups sparkling wine
54 1 67 55
specialty bar specialty cheese specialty chocolate specialty fat
269 84 299 36
specialty vegetables spices spread cheese sugar
17 51 110 333
sweet spreads syrup tea tidbits
89 32 38 23
toilet cleaner tropical fruit turkey UHT-milk
7 1032 80 329
vinegar waffles whipped/sour cream whisky
64 378 705 8
white bread white wine whole milk yogurt
414 187 2513 1372
zwieback
68Her bir items için absolute değelerini gözükmektedir.
Veri setine ait Absolute ve Relative grafiklerini çizdirelim:
Absolute Item Frequency Plot Top 25
Item Frequency Plot
jpeg(filename = "./Output/1-Absolute_ItemFrequencyPlot.jpg", width=1024, height=768, units="px")
itemFrequencyPlot(Groceries, topN=25, type="absolute",
main='Absolute Item Frequency Plot (Top 25)',
xlab="Items",
ylab="Item Frequency (Absolute)")
dev.off()
Relative Item Frequency Plot Top 25
jpeg(filename = "./Output/2-Relative_ItemFrequencyPlot.jpg", width=1024, height=768, units="px")
itemFrequencyPlot(Groceries, topN=25, type="relative",
main='Relative Item Frequency Plot (Top 25)',
xlab="Items",
ylab="Item Frequency (relative)")
dev.off()
Şimdi Birliktelik Kuralları oluşması için apriori modelini tasarlayalım;
Min. Support değerini %1 yani 0,01; Min. Confidence değerini %25 yani 0,25 olarak set ediyoruz. Ve Modelimizi bu parametreler ile initialize ediyoruz.
Not: Eğer Min. Support ve Min. Confidence değerlerini vermezisek; varsayılan başlangıç (%10 ve %80) değerlerini alarak model oluşturulacaktır.
# Both Association Rules
options(digits = 2)
min_supp <- 0.01
min_conf <- 0.25
min_lenght <- 2
BasketRules <- apriori(Groceries, parameter = list(supp = min_supp, conf = min_conf,
+ minlen = min_lenght, target = "rules"))
Apriori
Parameter specification:
confidence minval smax arem aval originalSupport maxtime support minlen maxlen target ext
0.25 0.1 1 none FALSE TRUE 5 0.01 2 10 rules FALSE
Algorithmic control:
filter tree heap memopt load sort verbose
0.1 TRUE TRUE FALSE TRUE 2 TRUE
Absolute minimum support count: 98
set item appearances ...[0 item(s)] done [0.00s].
set transactions ...[169 item(s), 9835 transaction(s)] done [0.07s].
sorting and recoding items ... [88 item(s)] done [0.00s].
creating transaction tree ... done [0.00s].
checking subsets of size 1 2 3 4 done [0.00s].
writing ... [170 rule(s)] done [0.00s].
creating S4 object ... done [0.00s].Verdiğimiz min support ve min confidence değelerince 170 adet rules oluştuğu görülmektedir.
summary(BasketRules)
set of 170 rules
rule length distribution (lhs + rhs):sizes
2 3
96 74
Min. 1st Qu. Median Mean 3rd Qu. Max.
2.000 2.000 2.000 2.435 3.000 3.000
summary of quality measures:
support confidence lift count
Min. :0.01007 Min. :0.2517 Min. :0.9932 Min. : 99.0
1st Qu.:0.01159 1st Qu.:0.2973 1st Qu.:1.5215 1st Qu.:114.0
Median :0.01454 Median :0.3587 Median :1.7784 Median :143.0
Mean :0.01822 Mean :0.3703 Mean :1.8747 Mean :179.2
3rd Qu.:0.02097 3rd Qu.:0.4253 3rd Qu.:2.1453 3rd Qu.:206.2
Max. :0.07483 Max. :0.5862 Max. :3.2950 Max. :736.0
mining info:
data ntransactions support confidence
Groceries 9835 0.01 0.25170 adet oluşan kuralların ilk 25 tanesini görelim;
# inspect(BasketRules)
inspect(BasketRules[1:25])
lhs rhs support confidence lift count
[1] {hard cheese} => {whole milk} 0.01006609 0.4107884 1.607682 99
[2] {butter milk} => {other vegetables} 0.01037112 0.3709091 1.916916 102
[3] {butter milk} => {whole milk} 0.01159126 0.4145455 1.622385 114
[4] {ham} => {whole milk} 0.01148958 0.4414062 1.727509 113
[5] {sliced cheese} => {whole milk} 0.01077783 0.4398340 1.721356 106
[6] {oil} => {whole milk} 0.01128622 0.4021739 1.573968 111
[7] {onions} => {other vegetables} 0.01423488 0.4590164 2.372268 140
[8] {onions} => {whole milk} 0.01209964 0.3901639 1.526965 119
[9] {berries} => {yogurt} 0.01057448 0.3180428 2.279848 104
[10] {berries} => {other vegetables} 0.01026945 0.3088685 1.596280 101
[11] {berries} => {whole milk} 0.01179461 0.3547401 1.388328 116
[12] {hamburger meat} => {other vegetables} 0.01382816 0.4159021 2.149447 136
[13] {hamburger meat} => {whole milk} 0.01474326 0.4434251 1.735410 145
[14] {hygiene articles} => {whole milk} 0.01281139 0.3888889 1.521975 126
[15] {salty snack} => {other vegetables} 0.01077783 0.2849462 1.472646 106
[16] {salty snack} => {whole milk} 0.01118454 0.2956989 1.157262 110
[17] {sugar} => {other vegetables} 0.01077783 0.3183183 1.645119 106
[18] {sugar} => {whole milk} 0.01504830 0.4444444 1.739400 148
[19] {waffles} => {other vegetables} 0.01006609 0.2619048 1.353565 99
[20] {waffles} => {whole milk} 0.01270971 0.3306878 1.294196 125
[21] {long life bakery product} => {other vegetables} 0.01067616 0.2853261 1.474610 105
[22] {long life bakery product} => {whole milk} 0.01352313 0.3614130 1.414444 133
[23] {dessert} => {other vegetables} 0.01159126 0.3123288 1.614164 114
[24] {dessert} => {whole milk} 0.01372649 0.3698630 1.447514 135
[25] {cream cheese} => {yogurt} 0.01240468 0.3128205 2.242412 122
>
> inspect(head(BasketRules, n=5))
lhs rhs support confidence lift count
[1] {hard cheese} => {whole milk} 0.01006609 0.4107884 1.607682 99
[2] {butter milk} => {other vegetables} 0.01037112 0.3709091 1.916916 102
[3] {butter milk} => {whole milk} 0.01159126 0.4145455 1.622385 114
[4] {ham} => {whole milk} 0.01148958 0.4414062 1.727509 113
[5] {sliced cheese} => {whole milk} 0.01077783 0.4398340 1.721356 106
> inspect(tail(BasketRules, n=5))
lhs rhs support confidence lift count
[1] {other vegetables,yogurt} => {whole milk} 0.02226741 0.5128806 2.007235 219
[2] {whole milk,yogurt} => {other vegetables} 0.02226741 0.3974592 2.054131 219
[3] {other vegetables,whole milk} => {yogurt} 0.02226741 0.2975543 2.132979 219
[4] {other vegetables,rolls/buns} => {whole milk} 0.01789527 0.4200477 1.643919 176
[5] {rolls/buns,whole milk} => {other vegetables} 0.01789527 0.3159785 1.633026 176Burada ilk kuralı yorumlayacak olursak;
lhs rhs support confidence lift count
[1] {hard cheese} => {whole milk} 0.01006609 0.4107884 1.607682 99Kuralın Sol Tarafını baz alarak başlıyoruz;
Support Yorumu: Bir fiş hareketi içerisinde Hard Cheese ürünü ile Whole Milk 0,01 yani %1 değeri ile birlikte oldukça az görülmektedir.
Confidence Yorumu: Aynı zamanda Hard Cheese ürününü satın alan bir kişi 0,41 yani %41 ile Whole Milk ürünün de satın alacağı görülmektedir. Kuralın RHS(Right Hand Side) yani Consequent tarafı ile ilgilenir.
Lift Yorumu: Hard Cheese ürününü olan alışveriş sepetlerinde Whole Milk ürünlerini satışı 1,60 kat artırıyor. Bu değer kuralın LHS(Left Hand Side) yani Antecedent tarafı ile ilgilenir.
Count Yorumu: Her iki items’ın birlikte görülme sıklığı 99 olduğu görülmüştür.
Şimdi de oluşturulan Apriori Modeline ait bilgileri ekrana yazdırıyoruz.
cat(paste0("\n","Association Rules Analysis", "\n",
+ "= = = = = = = = = = = = = =", "\n",
+ "Min. Supp. & Conf. Values:", "\t", min_supp, " & ", min_conf, "\n",
+ "X ==> Y Rules Count:", "\t\t", length(BasketRules)))
Association Rules Analysis
= = = = = = = = = = = = = =
Min. Supp. & Conf. Values: 0.01 & 0.25
X ==> Y Rules Count: 170Apriori algoritması kullanılarak Min. Supp ve Min. Conf değerleri ile 170 adet kural oluştuğu görülmektedir. Oluşan bu kuralları Condifence, Support ve Lift değerlerine göre descent(azalan) olarak sıralıyoruz. Ve devamında ilk ve son 5 adet kuralları listeliyoruz.
Confidence değerlerine göre ilk ve son 5 kural:
BasketRules_Conf <- sort(BasketRules, by="confidence", decreasing=TRUE)
inspect(head(BasketRules_Conf, n=5))
lhs rhs support confidence lift count
[1] {citrus fruit,root vegetables} => {other vegetables} 0.01037112 0.5862069 3.029608 102
[2] {root vegetables,tropical fruit} => {other vegetables} 0.01230300 0.5845411 3.020999 121
[3] {curd,yogurt} => {whole milk} 0.01006609 0.5823529 2.279125 99
[4] {butter,other vegetables} => {whole milk} 0.01148958 0.5736041 2.244885 113
[5] {root vegetables,tropical fruit} => {whole milk} 0.01199797 0.5700483 2.230969 118
inspect(tail(BasketRules_Conf, n=5))
lhs rhs support confidence lift count
[1] {fruit/vegetable juice} => {soda} 0.01840366 0.2545710 1.4598869 181
[2] {bottled beer} => {whole milk} 0.02043721 0.2537879 0.9932367 201
[3] {pastry} => {other vegetables} 0.02257245 0.2537143 1.3112349 222
[4] {chicken} => {root vegetables} 0.01087951 0.2535545 2.3262206 107
[5] {margarine} => {rolls/buns} 0.01474326 0.2517361 1.3686151 145Support değerlerine göre ilk ve son 5 kural:
BasketRules_Supp <- sort(BasketRules, by="support", decreasing=TRUE)
inspect(head(BasketRules_Supp, n=5))
lhs rhs support confidence lift count
[1] {other vegetables} => {whole milk} 0.07483477 0.3867578 1.513634 736
[2] {whole milk} => {other vegetables} 0.07483477 0.2928770 1.513634 736
[3] {rolls/buns} => {whole milk} 0.05663447 0.3079049 1.205032 557
[4] {yogurt} => {whole milk} 0.05602440 0.4016035 1.571735 551
[5] {root vegetables} => {whole milk} 0.04890696 0.4486940 1.756031 481
inspect(tail(BasketRules_Supp, n=5))
lhs rhs support confidence lift count
[1] {sausage,whole milk} => {other vegetables} 0.01016777 0.3401361 1.757876 100
[2] {hard cheese} => {whole milk} 0.01006609 0.4107884 1.607682 99
[3] {waffles} => {other vegetables} 0.01006609 0.2619048 1.353565 99
[4] {curd,yogurt} => {whole milk} 0.01006609 0.5823529 2.279125 99
[5] {curd,whole milk} => {yogurt} 0.01006609 0.3852140 2.761356 99Lift değerlerine göre ilk ve son 5 kural:
BasketRules_Lift <- sort(BasketRules, by="lift", decreasing=TRUE)
inspect(head(BasketRules_Lift, n=5))
lhs rhs support confidence lift count
[1] {citrus fruit,other vegetables} => {root vegetables} 0.01037112 0.3591549 3.295045 102
[2] {other vegetables,tropical fruit} => {root vegetables} 0.01230300 0.3427762 3.144780 121
[3] {beef} => {root vegetables} 0.01738688 0.3313953 3.040367 171
[4] {citrus fruit,root vegetables} => {other vegetables} 0.01037112 0.5862069 3.029608 102
[5] {root vegetables,tropical fruit} => {other vegetables} 0.01230300 0.5845411 3.020999 121
inspect(tail(BasketRules_Lift, n=5))
lhs rhs support confidence lift count
[1] {sausage} => {whole milk} 0.02989324 0.3181818 1.2452520 294
[2] {bottled water} => {whole milk} 0.03436706 0.3109476 1.2169396 338
[3] {rolls/buns} => {whole milk} 0.05663447 0.3079049 1.2050318 557
[4] {salty snack} => {whole milk} 0.01118454 0.2956989 1.1572618 110
[5] {bottled beer} => {whole milk} 0.02043721 0.2537879 0.9932367 201Şimdi ise oluşan kuralları visualise(görselleştirme) ederek somutlaştıralım. Bunun için Grap.R dosyasının içerisinde yazmış olduğum kod;
source('Graph.R')
cat("Export Graph for BasketRules")
jpeg(filename = "./Output/3-BasketRules_ScatterPlot.jpg", width=1024, height=768, units="px")
plot(BasketRules, main="Scatter Plot for Association TelcoRules")
dev.off()
jpeg(filename = "./Output/4-BasketRules_GroupedPlot.jpg", width=1024, height=768, units="px") plot(BasketRules, method="grouped", main="Grouped Matrix for Groceries") dev.off()
jpeg(filename = "./Output/5-BasketRules_GraphPlot.jpg", width=1024, height=768, units="px") plot(BasketRules[1:25], method="graph", control=list(alpha="1"), main="Graph Method for Groceries") dev.off()
jpeg(filename = "./Output/6-BasketRules_Paracoord.jpg", width=1024, height=768, units="px") plot(BasketRules, method ="paracoord", control = list(reorder = TRUE), main="GraphName6") dev.off()
Şimdi de Interactive Graph(İnteraktif Grafik) ile daha kullanışlı ve akıllı görselleştirmeler kullanalım. Bunun için InteractiveGraph.R dosyasının içerisinde yazmış olduğum kod;
# Export Interactive Graph
source('InteractiveGraph.R')cat("Interactive Graph")
if (file.exists("./Output/7-InteractiveGraph/") == FALSE){
dir.create(file.path("./Output/7-InteractiveGraph/"))
}
ip1 <- plotly_arules(BasketRules, measure = c("support", "confidence"), shading = "lift")
saveWidget(ip1, file = paste0(Path, "/Output/7-InteractiveGraph/", "InteractiveGraph-1.html"), selfcontained = FALSE)
ip2 <- plotly_arules(BasketRules, measure = c("support", "lift"), shading = "confidence")
saveWidget(ip2, file = paste0(Path, "/Output/7-InteractiveGraph/", "InteractiveGraph-2.html"), selfcontained = FALSE)
ip3 <- plotly_arules(BasketRules, method = "two-key plot", measure = c("support", "confidence"))
saveWidget(ip3, file = paste0(Path, "/Output/7-InteractiveGraph/", "InteractiveGraph-3.html"), selfcontained = FALSE)
ip4 <- plotly_arules(BasketRules, method = "two-key plot", measure = c("support", "lift"))
saveWidget(ip4, file = paste0(Path, "/Output/7-InteractiveGraph/", "InteractiveGraph-4.html"), selfcontained = FALSE)
ip5 <- plotly_arules(BasketRules, method = "matrix", measure = c("support", "lift"), shading = "confidence")
saveWidget(ip5, file = paste0(Path, "/Output/7-InteractiveGraph/", "InteractiveGraph-5.html"), selfcontained = FALSE)
ip6 <- plotly_arules(BasketRules, method = "matrix", measure = c("support", "confidence"), shading = "lift")
saveWidget(ip6, file = paste0(Path, "/Output/7-InteractiveGraph/", "InteractiveGraph-6.html"), selfcontained = FALSE)Bazı interaktif görselleri aşağıda paylaşıyorum. İnteraktif görselleri html dosyası olarak kaydedip. Herhangi bir tarayıcı ile açıp ilgili kuralların üzerine gelip bilgiler elde edebilirsiniz. Ne demek istediğimi Association Rules Analysis with R Github Repositories’ini incelediğinizde daha iyi anlayacaksınız.
Not: Interactive Graph çıktılarını Output dizini altında 7-InteractiveGraph dizinini içerisinde bulabilirsiniz.
Son olarak da Other Graph Format’ları olan (edgelist, graphml, gml, dot) kullanalım. Bunun için OtherGraph.R dosyasının içerisinde yazmış olduğum kod;
# Export Other Graph Format: edgelist, graphml, gml, dot
source('OtherGraph.R')cat("Export Other Format: edgelist, graphml, gml, dot")
Export Other Format: edgelist, graphml, gml, dot
if (file.exists("./Output/8-OtherGraphFormat") == FALSE){
dir.create(file.path("./Output/8-OtherGraphFormat"))
}
Tip <- c("items", "itemsets")
Format <- c("edgelist", "graphml", "dimacs", "gml", "dot")
for (t in Tip){
for (f in Format){
if (f == "dimacs"){
next
}
saveAsGraph(BasketRules, paste0("./Output/8-OtherGraphFormat/BasketRules_", t, ".", f), type=t, format=f)
}
}Not: Other Graph Formatları Items ve Itemset olarak Output dizini altında 8-OtherGraphFormat dizininde bulabilirsiniz.
# Export The Rules BasketRules_DF <- as(BasketRules, "data.frame") write.csv2(BasketRules_DF, file = "./Output/BasketRules.csv", row.names = FALSE)
Oluşan 170 adet kuralı Output dizini altına BasketRules.csv dosyasına kaydediyoruz.
# Export Interactive HTML The Rules
html_page <- inspectDT(BasketRules)
saveWidget(html_page, file = paste0(Path, "/Output/", "BasketRules.html"), selfcontained = FALSE)
# inspectDT(BasketRules)
# browseURL(paste0(Path, "/Output/", "BasketRules.html"), browser = getOption("browser"), encodeIfNeeded = FALSE)
Oluşan 170 adet kuralı Interactive bir formatta Output dizini altına BasketRules.html dosyasına kaydediyoruz. Çıktısı:
Modeli ve diğer tüm değişkenleri .RData olarak Output dizini içerisinde BasketRules.RData dosyasına kaydediyoruz.
save.image(file = "./Output/BasketRules.RData")
Yukarıdaki tüm kodlar;
# Association Rules Analysis with R Project ----------------------------
# Clean the System & Console Variable --------------------
rm(list = ls())
cat("\014")
options(warn = -1)
# Sys_Date <- format(Sys.Date(), "%Y%m%d")
# Sys_Time <- format(Sys.time(), "%H:%M:%S")
# Install & Use Library ---------------------------------------------------
if (require('arules') == FALSE){
install.packages('arules')
library(arules)
}
if (require('arulesViz') == FALSE){
install.packages('arulesViz')
library(arulesViz)
}
if (require('htmlwidgets') == FALSE){
install.packages('htmlwidgets')
library(htmlwidgets)
}
# Set the System Path & Variable
if (require('here') == FALSE){
install.packages('here')
library(here)
}
Path <- here()
setwd(Path)
# Load Dataset
Groceries = read.transactions(file = "./Input/groceries.csv",
format = c("basket"),
sep= ",",
cols = NULL,
rm.duplicates = TRUE,
encoding = "UTF-8")
# Second Way Data Load
# data(Groceries)
# Show Type and Class of Dataset
typeof(Groceries)
class(Groceries)
# Dimension of Dataset
dim(Groceries)
length(Groceries)
# Structure of Dataset
str(Groceries)
# Show The Dataset
# inspect(Groceries)
inspect(Groceries[1:10])
# Show the Items
ItemSetList <- Groceries@itemInfo
ItemSetList
# First 5 items of DataSet
head(ItemSetList, n=10)
head(Groceries@itemInfo, n=10)
# End 5 ites of DataSet
tail(ItemSetList, n=10)
tail(Groceries@itemInfo, n=10)
# Summary of Dataset
summary(Groceries)
s <- summary(Groceries)
# Most Frequent Items
s@itemSummary
# Length Distribution Size
s@lengths
# frequency/support
itemFrequency(Groceries, type = "relative")
# frequency/support (Items Count)
itemFrequency(Groceries, type = "absolute")
# Item Frequency Plot
jpeg(filename = "./Output/1-Absolute_ItemFrequencyPlot.jpg", width=1024, height=768, units="px")
itemFrequencyPlot(Groceries, topN=25, type="absolute",
main='Absolute Item Frequency Plot (Top 25)',
xlab="Items",
ylab="Item Frequency (Absolute)")
dev.off()
jpeg(filename = "./Output/2-Relative_ItemFrequencyPlot.jpg", width=1024, height=768, units="px")
itemFrequencyPlot(Groceries, topN=25, type="relative",
main='Relative Item Frequency Plot (Top 25)',
xlab="Items",
ylab="Item Frequency (relative)")
dev.off()
# Both Association Rules
options(digits = 2)
min_supp <- 0.01
min_conf <- 0.25
min_lenght <- 2
BasketRules <- apriori(Groceries, parameter = list(supp = min_supp, conf = min_conf,
minlen = min_lenght, target = "rules"))
summary(BasketRules)
#inspect(BasketRules)
inspect(BasketRules[1:25])
inspect(head(BasketRules, n=5))
inspect(tail(BasketRules, n=5))
cat(paste0("\n","Association Rules Analysis", "\n",
"= = = = = = = = = = = = = =", "\n",
"Min. Supp. & Conf. Values:", "\t", min_supp, " & ", min_conf, "\n",
"X ==> Y Rules Count:", "\t\t", length(BasketRules)))
BasketRules_Conf <- sort(BasketRules, by="confidence", decreasing=TRUE)
BasketRules_Supp <- sort(BasketRules, by="support", decreasing=TRUE)
BasketRules_Lift <- sort(BasketRules, by="lift", decreasing=TRUE)
inspect(head(BasketRules_Conf, n=5))
inspect(tail(BasketRules_Conf, n=5))
inspect(head(BasketRules_Supp, n=5))
inspect(tail(BasketRules_Supp, n=5))
inspect(head(BasketRules_Lift, n=5))
inspect(tail(BasketRules_Lift, n=5))
# Export Graph
source('Graph.R')
cat("Export Graph for BasketRules")
jpeg(filename = "./Output/3-BasketRules_ScatterPlot.jpg", width=1024, height=768, units="px")
plot(BasketRules, main="Scatter Plot for Association TelcoRules")
dev.off()
jpeg(filename = "./Output/4-BasketRules_GroupedPlot.jpg", width=1024, height=768, units="px")
plot(BasketRules, method="grouped", main="Grouped Matrix for Groceries")
dev.off()
jpeg(filename = "./Output/5-BasketRules_GraphPlot.jpg", width=1024, height=768, units="px")
plot(BasketRules[1:25], method="graph", control=list(alpha="1"), main="Graph Method for Groceries")
dev.off()
jpeg(filename = "./Output/6-BasketRules_Paracoord.jpg", width=1024, height=768, units="px")
plot(BasketRules, method ="paracoord", control = list(reorder = TRUE), main="GraphName6")
dev.off()
# Export Interactive Graph
source('InteractiveGraph.R')
cat("Interactive Graph")
if (file.exists("./Output/7-InteractiveGraph/") == FALSE){
dir.create(file.path("./Output/7-InteractiveGraph/"))
}
ip1 <- plotly_arules(BasketRules, measure = c("support", "confidence"), shading = "lift")
saveWidget(ip1, file = paste0(Path, "/Output/7-InteractiveGraph/", "InteractiveGraph-1.html"), selfcontained = FALSE)
ip2 <- plotly_arules(BasketRules, measure = c("support", "lift"), shading = "confidence")
saveWidget(ip2, file = paste0(Path, "/Output/7-InteractiveGraph/", "InteractiveGraph-2.html"), selfcontained = FALSE)
ip3 <- plotly_arules(BasketRules, method = "two-key plot", measure = c("support", "confidence"))
saveWidget(ip3, file = paste0(Path, "/Output/7-InteractiveGraph/", "InteractiveGraph-3.html"), selfcontained = FALSE)
ip4 <- plotly_arules(BasketRules, method = "two-key plot", measure = c("support", "lift"))
saveWidget(ip4, file = paste0(Path, "/Output/7-InteractiveGraph/", "InteractiveGraph-4.html"), selfcontained = FALSE)
ip5 <- plotly_arules(BasketRules, method = "matrix", measure = c("support", "lift"), shading = "confidence")
saveWidget(ip5, file = paste0(Path, "/Output/7-InteractiveGraph/", "InteractiveGraph-5.html"), selfcontained = FALSE)
ip6 <- plotly_arules(BasketRules, method = "matrix", measure = c("support", "confidence"), shading = "lift")
saveWidget(ip6, file = paste0(Path, "/Output/7-InteractiveGraph/", "InteractiveGraph-6.html"), selfcontained = FALSE)
# Export Other Graph Format: edgelist, graphml, gml, dot
source('OtherGraph.R')
cat("Export Other Format: edgelist, graphml, gml, dot")
if (file.exists("./Output/8-OtherGraphFormat") == FALSE){
dir.create(file.path("./Output/8-OtherGraphFormat"))
}
Tip <- c("items", "itemsets")
Format <- c("edgelist", "graphml", "dimacs", "gml", "dot")
for (t in Tip){
for (f in Format){
if (f == "dimacs"){
next
}
saveAsGraph(BasketRules, paste0("./Output/8-OtherGraphFormat/BasketRules_", t, ".", f), type=t, format=f)
}
}
# Export The Rules
BasketRules_DF <- as(BasketRules, "data.frame")
write.csv2(BasketRules_DF, file = "./Output/BasketRules.csv", row.names = FALSE)
# Export Interactive HTML The Rules
html_page <- inspectDT(BasketRules)
saveWidget(html_page, file = paste0(Path, "/Output/", "BasketRules.html"), selfcontained = FALSE)
# inspectDT(BasketRules)
# browseURL(paste0(Path, "/Output/", "BasketRules.html"), browser = getOption("browser"), encodeIfNeeded = FALSE)
save.image(file = "./Output/BasketRules.RData")Son Olarak; Github üzerinde yer alan repository’e buradan (kod, input, output ve diğer dosyalara) ulaşabilirsiniz. Yazıyı beğenerek ve paylaşarak daha fazla kişiye ulaşmasına yardımcı olabilirsiniz. Görüş ve önerileriniz için uslumetin@gmail.com’dan bana ulaşabilirsiniz.
Yeniden görüşmek üzere, Selamlar 🙂
Bir Yanıt
Bir yanıt yazın
Yorum yapabilmek için oturum açmalısınız.
merhabalar, arulesViz paketi yeni versiyonu sanırım sıkıntı çıkarıyor, hata veriyor plotly_arules fonksiyonu paketten çıkarılmış gibi.