Assignment.Rmd

---
title: "Cafe Great dataset analysis"
author: "Gaurang Swarge"
date: "7/10/2018"
output: html_document
---

```{r setup}
knitr::opts_chunk$set(fig.width=12, fig.height=8, fig.path='Figs/', echo=FALSE, warning=FALSE, message=FALSE)

options(width = 100)

library(plyr)
library(data.table)
library(dplyr)
library(ggplot2)
library(stringr)
library(DT)
library(scales)
library(knitr)
library(plotly)
library(arules)
library(arulesViz)
library(visNetwork)
library(igraph)
library(kableExtra)
library(tidyverse)
library("RColorBrewer")

#Reading all the Datasets
orders <- fread("Cafe Great Transaction Data set.csv")
orders$`Item Desc` <- as.factor(orders$`Item Desc`)
orders$Category <- as.factor(orders$Category)
orders$date <- as.Date(as.character(orders$date),format="%dd-%mmm-%yy")
head(orders)
```

### Price Analysis
```{r PriceAnalysis}
summary(orders)

temp <- orders %>%
  filter(orders)

```

## Association Rules

```{r AssociationRules}


data1 <- split(orders$`Item Desc`,orders$`Bill Number`)

head(data1,25)

#converting to class transactions
total_transactions<-as(data1,"transactions")
dim(total_transactions)
inspect(total_transactions[1:2])

# calculates support for frequent items
frequentItems <- eclat (total_transactions, parameter = list(supp = 0.01, maxlen = 15)) 
inspect(frequentItems)

#frequent Items in the transactions
itemFrequencyPlot(total_transactions,topN=30,type="absolute")

#Running the Apriori algorithm
basket_rules <- apriori(total_transactions, parameter = list(support = 0.001, confidence = 0.10, minlen=2,maxlen=10,target = "rules"))
basket_rules

#Removing Redundent Rules (rules that are subset of larger rules)
# get subset rules in vector
subsetRules <- which(colSums(is.subset(basket_rules, basket_rules)) > 1) 
length(subsetRules) 
# remove subset rules 
basket_rules <- basket_rules[-subsetRules]
basket_rules
# 'high-confidence' rules.
basket_rules_conf <- sort (basket_rules, by="confidence", decreasing=TRUE) 

# show the support, lift and confidence for all rules
inspect(head(basket_rules_conf)) 

# High-Lift Rules
basket_rules_lift <- sort (basket_rules, by="lift", decreasing=TRUE) 

# show the support, lift and confidence for all rules
inspect(head(basket_rules_lift)) 

plot(basket_rules,control=list(col=brewer.pal(11,"Spectral")),main="")
plot(basket_rules, method = "graph")
plotly_arules(basket_rules)
inspectDT(basket_rules)

```