Association Analysis

---

Part Assocication Analysis

Part a

install.packages("arules")

Error in install.packages : Updating loaded packages

library(arules)
library(arulesViz)
tr1 <- read.transactions("/Users/jayavarshini/Desktop/ms/sem5/dm/assing2/tr-1k-canonical.csv", sep=",")
tr5 <- read.transactions("/Users/jayavarshini/Desktop/ms/sem5/dm/assing2/tr-5k-canonical.csv", sep=",")
tr20 <- read.transactions("/Users/jayavarshini/Desktop/ms/sem5/dm/assing2/tr-20k-canonical.csv", sep=",")

Restarting R session...

summary(itemFrequency(tr1))

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
0.02600 0.05675 0.07550 0.07076 0.08400 0.10800 

summary(itemFrequency(tr5))

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
0.03820 0.06250 0.07530 0.07094 0.08325 0.11080 

summary(itemFrequency(tr20))

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
0.04055 0.06294 0.07365 0.07113 0.08420 0.10985 

summary(itemFrequency(tr75))

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
0.04193 0.06211 0.07511 0.07099 0.08344 0.10924 

f_is <- apriori(tr1, parameter=list(support=0.045, target="frequent itemsets"))

Apriori

Parameter specification:

Algorithmic control:

Absolute minimum support count: 45 

set item appearances ...[0 item(s)] done [0.00s].
set transactions ...[50 item(s), 1000 transaction(s)] done [0.00s].
sorting and recoding items ... [42 item(s)] done [0.00s].
creating transaction tree ... done [0.00s].
checking subsets of size 1 2 done [0.00s].
writing ... [47 set(s)] done [0.00s].
creating S4 object  ... done [0.00s].

inspect(sort(f_is, decreasing = T , by="count"))

rm(f_is)
itemFrequencyPlot(tr1)

f_is <- apriori(tr5, parameter=list(support=0.05, target="frequent itemsets"))

Apriori

Parameter specification:

Algorithmic control:

Absolute minimum support count: 250 

set item appearances ...[0 item(s)] done [0.00s].
set transactions ...[50 item(s), 5000 transaction(s)] done [0.00s].
sorting and recoding items ... [38 item(s)] done [0.00s].
creating transaction tree ... done [0.00s].
checking subsets of size 1 2 done [0.00s].
writing ... [39 set(s)] done [0.00s].
creating S4 object  ... done [0.00s].

inspect(sort(f_is, decreasing = T , by="count"))

rm(f_is)
itemFrequencyPlot(tr5)

f_is <- apriori(tr20, parameter=list(support=0.06, target="frequent itemsets"))

Apriori

Parameter specification:

Algorithmic control:

Absolute minimum support count: 1200 

set item appearances ...[0 item(s)] done [0.00s].
set transactions ...[50 item(s), 20000 transaction(s)] done [0.01s].
sorting and recoding items ... [38 item(s)] done [0.00s].
creating transaction tree ... done [0.02s].
checking subsets of size 1 2 done [0.00s].
writing ... [38 set(s)] done [0.00s].
creating S4 object  ... done [0.01s].

inspect(sort(f_is, decreasing = T , by="count"))

rm(f_is)
itemFrequencyPlot(tr20)

f_is <- apriori(tr75, parameter=list(support=0.045, target="frequent itemsets"))

Apriori

Parameter specification:

Algorithmic control:

Absolute minimum support count: 3375 

set item appearances ...[0 item(s)] done [0.00s].
set transactions ...[50 item(s), 75000 transaction(s)] done [0.05s].
sorting and recoding items ... [38 item(s)] done [0.01s].
creating transaction tree ... done [0.08s].
checking subsets of size 1 2 done [0.00s].
writing ... [40 set(s)] done [0.00s].
creating S4 object  ... done [0.03s].

inspect(sort(f_is, decreasing = T , by="count"))

rm(f_is)
itemFrequencyPlot(tr75)

Using a support value of 0.5 gives us 2 item sets and individual item sets with a good number of record anf the most frequent item set Gongolias Cookie occurs 108 time with a good support value

r <- apriori(tr1,parameter = list(support=0.009))

Apriori

Parameter specification:

Algorithmic control:

Absolute minimum support count: 9 

set item appearances ...[0 item(s)] done [0.00s].
set transactions ...[50 item(s), 1000 transaction(s)] done [0.00s].
sorting and recoding items ... [50 item(s)] done [0.00s].
creating transaction tree ... done [0.00s].
checking subsets of size 1 2 3 4 5 done [0.00s].
writing ... [84 rule(s)] done [0.00s].
creating S4 object  ... done [0.00s].

summary(r)

set of 84 rules

rule length distribution (lhs + rhs):sizes
 3  4  5 
57 23  4 

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
  3.000   3.000   3.000   3.369   4.000   5.000 

summary of quality measures:
    support          confidence          lift            count      
 Min.   :0.01800   Min.   :0.8000   Min.   : 8.511   Min.   :18.00  
 1st Qu.:0.01900   1st Qu.:0.9044   1st Qu.:11.218   1st Qu.:19.00  
 Median :0.02400   Median :0.9554   Median :12.545   Median :24.00  
 Mean   :0.02568   Mean   :0.9410   Mean   :12.644   Mean   :25.68  
 3rd Qu.:0.02950   3rd Qu.:1.0000   3rd Qu.:14.124   3rd Qu.:29.50  
 Max.   :0.04000   Max.   :1.0000   Max.   :19.608   Max.   :40.00  

mining info:

inspect(head(r,by="confidence"))

r1 <- apriori(tr1,parameter = list(support=0.005))

Apriori

Parameter specification:

Algorithmic control:

Absolute minimum support count: 5 

set item appearances ...[0 item(s)] done [0.00s].
set transactions ...[50 item(s), 1000 transaction(s)] done [0.00s].
sorting and recoding items ... [50 item(s)] done [0.00s].
creating transaction tree ... done [0.00s].
checking subsets of size 1 2 3 4 5 done [0.00s].
writing ... [88 rule(s)] done [0.00s].
creating S4 object  ... done [0.00s].

summary(r1)

set of 88 rules

rule length distribution (lhs + rhs):sizes
 3  4  5 
61 23  4 

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
  3.000   3.000   3.000   3.352   4.000   5.000 

summary of quality measures:
    support          confidence          lift            count      
 Min.   :0.00500   Min.   :0.8000   Min.   : 8.403   Min.   : 5.00  
 1st Qu.:0.01900   1st Qu.:0.9032   1st Qu.:11.029   1st Qu.:19.00  
 Median :0.02400   Median :0.9500   Median :12.516   Median :24.00  
 Mean   :0.02476   Mean   :0.9367   Mean   :12.490   Mean   :24.76  
 3rd Qu.:0.02900   3rd Qu.:1.0000   3rd Qu.:13.946   3rd Qu.:29.00  
 Max.   :0.04000   Max.   :1.0000   Max.   :19.608   Max.   :40.00  

mining info:

install.packages("arulesViz")

trying URL 'https://cran.rstudio.com/bin/macosx/el-capitan/contrib/3.5/arulesViz_1.3-2.tgz'
Content type 'application/x-gzip' length 762680 bytes (744 KB)
==================================================
downloaded 744 KB

The downloaded binary packages are in
    /var/folders/6f/tkdd1c4j4pbbkbbzp07y2rgw0000gn/T//RtmpviW8Jj/downloaded_packages

library(arulesViz)

Loading required package: grid

inspect(head(r1,by="confidence"))

plot(r1,engine="htmlwidget",jitter=0)

r20 <- apriori(tr20,parameter = list(support=0.005))

Apriori

Parameter specification:

Algorithmic control:

Absolute minimum support count: 100 

set item appearances ...[0 item(s)] done [0.00s].
set transactions ...[50 item(s), 20000 transaction(s)] done [0.01s].
sorting and recoding items ... [50 item(s)] done [0.00s].
creating transaction tree ... done [0.01s].
checking subsets of size 1 2 3 4 5 done [0.01s].
writing ... [81 rule(s)] done [0.00s].
creating S4 object  ... done [0.01s].

summary(r20)

set of 81 rules

rule length distribution (lhs + rhs):sizes
 3  4  5 
52 24  5 

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
   3.00    3.00    3.00    3.42    4.00    5.00 

summary of quality measures:
    support          confidence          lift            count      
 Min.   :0.02040   Min.   :0.8016   Min.   : 7.942   Min.   :408.0  
 1st Qu.:0.02055   1st Qu.:0.9128   1st Qu.:12.533   1st Qu.:411.0  
 Median :0.02545   Median :0.9225   Median :13.407   Median :509.0  
 Mean   :0.02471   Mean   :0.9380   Mean   :13.015   Mean   :494.1  
 3rd Qu.:0.02810   3rd Qu.:0.9927   3rd Qu.:14.294   3rd Qu.:562.0  
 Max.   :0.04100   Max.   :1.0000   Max.   :15.026   Max.   :820.0  

mining info:

inspect(head(r20,by="confidence"))

    lhs                     rhs                  support confidence     lift count
[1] {Green Tea,                                                                   
     Lemon Cookie,                                                                
     Lemon Lemonade,                                                              
     Raspberry Cookie}   => {Raspberry Lemonade} 0.02040  1.0000000 14.60920   408
[2] {Green Tea,                                                                   
     Lemon Cookie,                                                                
     Raspberry Cookie,                                                            
     Raspberry Lemonade} => {Lemon Lemonade}     0.02040  1.0000000 15.02630   408
[3] {Almond Twist,                                                                
     Coffee Eclair,                                                               
     Hot Coffee}         => {Apple Pie}          0.02810  0.9982238 13.46222   562
[4] {Green Tea,                                                                   
     Raspberry Cookie,                                                            
     Raspberry Lemonade} => {Lemon Lemonade}     0.02045  0.9975610 14.98965   409
[5] {Green Tea,                                                                   
     Lemon Cookie,                                                                
     Lemon Lemonade,                                                              
     Raspberry Lemonade} => {Raspberry Cookie}   0.02040  0.9975550 14.36364   408
[6] {Green Tea,                                                                   
     Lemon Lemonade,                                                              
     Raspberry Cookie,                                                            
     Raspberry Lemonade} => {Lemon Cookie}       0.02040  0.9975550 14.61619   408

plot(r20,engine="htmlwidget",jitter=0)

r5<- apriori(tr5,parameter = list(support=0.005))

Apriori

Parameter specification:

Algorithmic control:

Absolute minimum support count: 25 

set item appearances ...[0 item(s)] done [0.00s].
set transactions ...[50 item(s), 5000 transaction(s)] done [0.00s].
sorting and recoding items ... [50 item(s)] done [0.00s].
creating transaction tree ... done [0.00s].
checking subsets of size 1 2 3 4 5 done [0.00s].
writing ... [85 rule(s)] done [0.00s].
creating S4 object  ... done [0.00s].

summary(r5)

set of 85 rules

rule length distribution (lhs + rhs):sizes
 3  4  5 
53 27  5 

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
  3.000   3.000   3.000   3.435   4.000   5.000 

summary of quality measures:
    support          confidence          lift            count      
 Min.   :0.02120   Min.   :0.8030   Min.   : 7.859   Min.   :106.0  
 1st Qu.:0.02120   1st Qu.:0.9120   1st Qu.:12.059   1st Qu.:106.0  
 Median :0.02620   Median :0.9296   Median :13.478   Median :131.0  
 Mean   :0.02573   Mean   :0.9341   Mean   :13.149   Mean   :128.7  
 3rd Qu.:0.02980   3rd Qu.:0.9913   3rd Qu.:14.621   3rd Qu.:149.0  
 Max.   :0.04080   Max.   :1.0000   Max.   :15.625   Max.   :204.0  

mining info:

inspect(head(r5,by="confidence"))

plot(r5,engine="htmlwidget",jitter=0)

These rules have good confidence and high lift showing these are intresting findings.

r75 <- apriori(tr75, parameter = list(support=0.005))

Apriori

Parameter specification:

Algorithmic control:

Absolute minimum support count: 375 

set item appearances ...[0 item(s)] done [0.00s].
set transactions ...[50 item(s), 75000 transaction(s)] done [0.06s].
sorting and recoding items ... [50 item(s)] done [0.01s].
creating transaction tree ... done [0.07s].
checking subsets of size 1 2 3 4 5 done [0.01s].
writing ... [85 rule(s)] done [0.00s].
creating S4 object  ... done [0.04s].

summary(r75)

set of 85 rules

rule length distribution (lhs + rhs):sizes
 3  4  5 
52 28  5 

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
  3.000   3.000   3.000   3.447   4.000   5.000 

summary of quality measures:
    support          confidence          lift            count     
 Min.   :0.02059   Min.   :0.8049   Min.   : 7.924   Min.   :1544  
 1st Qu.:0.02073   1st Qu.:0.9070   1st Qu.:12.873   1st Qu.:1555  
 Median :0.02093   Median :0.9229   Median :13.403   Median :1570  
 Mean   :0.02454   Mean   :0.9319   Mean   :13.035   Mean   :1840  
 3rd Qu.:0.02676   3rd Qu.:0.9922   3rd Qu.:14.514   3rd Qu.:2007  
 Max.   :0.04111   Max.   :1.0000   Max.   :14.784   Max.   :3083  

mining info:

inspect(head(r75, by="confidence"))

    lhs                     rhs                     support confidence     lift count
[1] {Green Tea,                                                                      
     Lemon Cookie,                                                                   
     Raspberry Cookie,                                                               
     Raspberry Lemonade} => {Lemon Lemonade}     0.02073333  1.0000000 14.65416  1555
[2] {Green Tea,                                                                      
     Lemon Cookie,                                                                   
     Lemon Lemonade,                                                                 
     Raspberry Cookie}   => {Raspberry Lemonade} 0.02073333  1.0000000 14.76087  1555
[3] {Green Tea,                                                                      
     Lemon Cookie,                                                                   
     Lemon Lemonade,                                                                 
     Raspberry Lemonade} => {Raspberry Cookie}   0.02073333  1.0000000 14.78415  1555
[4] {Green Tea,                                                                      
     Lemon Lemonade,                                                                 
     Raspberry Cookie,                                                               
     Raspberry Lemonade} => {Lemon Cookie}       0.02073333  0.9993573 14.69355  1555
[5] {Green Tea,                                                                      
     Lemon Cookie,                                                                   
     Raspberry Cookie}   => {Raspberry Lemonade} 0.02073333  0.9980745 14.73245  1555
[6] {Green Tea,                                                                      
     Lemon Cookie,                                                                   
     Raspberry Cookie}   => {Lemon Lemonade}     0.02073333  0.9980745 14.62594  1555

plot(r75,engine="htmlwidget",jitter=0)

Part b

The results of the rules seem to have pretty good lift and confidence. Increasing the number of the number of transaction seem to have provided difference results with much more clarity. For example, the Consequent with Rasberry Lemonade is present with less support value with antecedent in 1k dataset while in the 75k dataset it is the best rule that is determined. The accuracy of the rules are increased as the number of data in the transcation is increased.

Part c

f_is <- apriori(tr75, parameter=list(support=0.05, target="frequent itemsets"))

Apriori

Parameter specification:

Algorithmic control:

Absolute minimum support count: 3750 

set item appearances ...[0 item(s)] done [0.00s].
set transactions ...[50 item(s), 75000 transaction(s)] done [0.06s].
sorting and recoding items ... [38 item(s)] done [0.01s].
creating transaction tree ... done [0.08s].
checking subsets of size 1 2 done [0.00s].
writing ... [40 set(s)] done [0.00s].
creating S4 object  ... done [0.04s].

inspect(sort(f_is, decreasing = T , by="count"))

rm(f_is)
itemFrequencyPlot(tr75)

1. Coffee Eclair seems to be purchased the most ii.{Marzipan Cookie,Tuile Cookie} seems to be purchased less for the support value I took.