---
title: "Public distribution system detect corruption of sugar using Kmeans algorithm"
output: 
  flexdashboard::flex_dashboard:
    orientation: columns
    vertical_layout: scroll
    theme: simplex
    social: menu
    source_code: embed
---

```{r setup, include=FALSE}
library(flexdashboard)
library(tidyverse)
library(cluster)
library(dplyr)
library(lattice)
library(readr)
library(dendextend)
```

Column {data-width=650}
-----------------------------------------------------------------------

```{r}
#Divisive clustering
#distance matrix
data <- read.csv("C:\\Users\\India\\Documents\\Dma\\shop-wise-trans-detailsTelangana State Civil Supplies Distribution - FP Shop Wise Transactions Data_01-01-2021 to 31-01-2021.csv")
distric_class = data.frame(
  distName=c("Adilabad","Bhadrdri Kothagudem","Hanumakonda","Hyderabad","Jagityal","Janagaon","Jayashankar Bhupalpalli","Jogulamba Gadwal","Kamareddy","Karimnagar","Khammam","Kumarambheem Asifabad","Mahabubabad","Mahbubnagar","Manchiryala","Medak","Medchal","Mulugu","Nagarkarnool","Nalgonda","Narayanpet","Nirmal","Nizamabad","Peddapalli","Rajanna Siricilla","Ranga Reddy","Sangareddy","Siddipet","Suryapet","Vikarabad","Wanaparthy","Warangal","Yadadri Bhuvanagiri" ),
  Classifi=c("Rural","Rural","Urban","Urban","Rural","Urban","Rural","Rural","Rural","Rural","Rural","Rural","Rural","Rural","Rural","Rural","Rural","Rural","Rural","Rural","Rural","Urban","Rural","Rural","Rural","Urban","Rural","Rural","Urban","Rural","Urban","Rural","Rural"))
#Merge the classification data with the main dataset based on district names
sub1=merge(data,distric_class,by='distName',,all.x=TRUE)
sub=data.frame(sub1$Classifi,sub1$distCode,sub1$officeCode,sub1$shopNo)

"Organization: Telangana State Civil Supplies Corporation Limited"

"The dataset about the FP Shop Distribution data in Telangana."

"This dataset contains 16970 obs. of  13 variables:"  
  
"distCode : District Code"

"officeCode : Represents the office code"

"shopNo   : Representing Shop number"

"noOfRcs  : No.of Ration Cards"

"noOfTrans: No.of Transactions"

"riceAfsc : Rice to Anthtodaya Food Security Cards"

"riceFsc  : Rice to Food Security Card"

"riceAap  : Rice to Annapurna Card"

"sugar    : sugar quantity"

"totalAmount : total amount"

"otherShopTransCnt : Other shops transaction quantity"

```


### Distance measure

```{r}
head(sub)
distancematrix=dist(sub,method='euclidean')
```


### plot of hierarchical view of data

```{r}
set.seed(40)
hier_clus=hclust(distancematrix,method='average')
plot(hier_clus,hang=0.1,main='hierarchical clustering',xlab = "",ylab="Height")
```

### dendrogram
```{r}
library(dendextend)
dend=as.dendrogram(hier_clus)
COLS=c("orange","turquoise","pink","lightblue","darkgreen","blue","green","white","brown","black","yellow","purple")
#names(COLS)=unique(sub1)
dend=color_labels(dend, col=COLS)
plot(dend,main='hierarchical clustering',xlab = "PDS levels ",ylab="Height")
rect.hclust(hier_clus,k=3,border = 'red')
```


### accuracy of clustering

```{r}
fit=cutree(hier_clus,k=3)
table(fit)
cm=as.matrix(table(fit))
cm
accuracy = sum(diag(cm))/sum(cm)
accuracy
```