R/univariate.R
In nolancardozo13/eda: Automated Exploratory Data Analysis
library(R6)
univariate <- R6Class(
"Univariate Analysis",
portable = FALSE,
cloneable = FALSE,
public = list(
columns = list(),
metadata = list(),
ngrams = list(),
initialize = function(metadata,k = 3) {
options(scipen = 999)
dist <- function(var){
a<-cumsum(as.numeric(var))
b<- a/sum(as.numeric(var))
c<- length(which(b<=0.8))
d<- c/length(var)
return(d)
}
findNGrams <- function(df, n) {
ngram <- NGramTokenizer(df, Weka_control(min = n, max = n, delimiters = " \\r\\n\\t.,;:\"()?!"))
ngram2 <- data.frame(table(ngram))
ngram3 <- ngram2[order(ngram2$Freq, decreasing = TRUE),]
colnames(ngram3) <- c("Text","Count")
head(ngram3,10)
}
if(!missing(metadata))
{
self$metadata <- as.list.environment(metadata)
metadata <- as.list.environment(metadata)
data <- metadata$data
metadata1 <- NULL
for(i in 1:length(metadata$columns)){
metadata1 <- rbind(metadata1,as.data.frame(metadata$columns[[i]]))
}
num_var <- as.character(metadata1[which(metadata1$type %in% c("discrete","continuous")),"column_name"])
cat_var <- as.character(metadata1[which(metadata1$type %in% c("character","ordered")),"column_name"])
text_var <- as.character(metadata1[which(metadata1$type %in% c("text")),"column_name"])
self$ngrams <- vector("list", length(text_var))
names(self$ngrams) <- text_var
self$columns <- vector("list", length(num_var))
names(self$columns) <- num_var
for(q in 1:length(text_var)){
self$ngrams[[q]] <- vector("list",4)
names(self$ngrams[[q]]) <- c("TwoGram","ThreeGram","FourGram","FiveGram")
}
mydataframe <- c("column_name","upper_outliers_IQR","lower_outliers_IQR",paste("upper_outliers_",k,"sigma",sep = ""),paste("lower_outliers_",k,"sigma",sep = ""),"concentration","priority","performance","notes")
for(i in 1:ncol(data)){
if(names(data)[i] %in% num_var){
self$columns[[i]] <- setNames(data.frame(matrix(ncol = length(mydataframe), nrow = 0)), mydataframe)
a1 <- names(data)[i]
q1 <-quantile(data[,i],0.25,na.rm = T)
q3 <- quantile(data[,i],0.75,na.rm =T )
iqr <- q3 - q1
a2 <- q3 + 1.5*iqr
a3 <- q1 - 1.5*iqr
a4 <- paste(round(((length(which(data[,i] > a2))*100)/(nrow(data))),1),"%")
a5 <- paste(round(((length(which(data[,i] < a3))*100)/(nrow(data))),1),"%")
a6 <- mean(data[,i],na.rm = T) + k*sd(data[,i],na.rm = T)
a7 <- mean(data[,i],na.rm = T) - k*sd(data[,i],na.rm = T)
a8 <- paste(round(((length(which(data[,i] > a6))*100)/(nrow(data))),1),"%")
a9 <- paste(round(((length(which(data[,i] < a7))*100)/(nrow(data))),1),"%")
a13 <- paste(round((dist(data[,i]) *100)),"%")
a10 <- NA
a11 <- NA
a12 <- NA
self$columns[[i]] <- list("a1" = a1,"a4"=a4,"a5"=a5,"a8"=a8,"a9"=a9,"a10" = a13,"a11"= a10,"a12" = a11,"a13" = a12)
names(self$columns[[i]]) <- mydataframe
rm(a1); rm(a2); rm(a3); rm(a4); rm(a5); rm(a6); rm(a7); rm(a8); rm(a9); rm(a10); rm(a11); rm(a12); rm(a13);
}
else if(names(data)[i] %in% text_var){
for(w in 1:length(text_var)){
self$ngrams[[w]]$TwoGram <- findNGrams(data[,i], 2)
self$ngrams[[w]]$ThreeGram <- findNGrams(data[,i], 3)
self$ngrams[[w]]$FourGram <- findNGrams(data[,i], 4)
self$ngrams[[w]]$FiveGram <- findNGrams(data[,i], 5)
}
}
}
}
},
save = function(savepath,sheet= "Univariate"){
require(xlsx)
unidata <- NULL
for(i in 1:length(self$columns)){
unidata <- rbind(unidata,as.data.frame(self$columns[[i]]))
}
names(unidata) <- gsub("_"," ",names(unidata))
if(!file.exists(savepath)){
wb1 = createWorkbook()
sheet = createSheet(wb1,sheet)
ngramsheet = createSheet(wb1,"Ngrams")
csTableColNames <- CellStyle(wb1) + Font(wb1, isBold=TRUE) + Alignment(wrapText=TRUE, h="ALIGN_CENTER") + Border(color="black", position=c("TOP", "BOTTOM"), pen=c("BORDER_THIN", "BORDER_THICK"))
addDataFrame(unidata, sheet=sheet,startRow=2,startColumn=1, row.names=FALSE,colnamesStyle=csTableColNames)
setColumnWidth(sheet, colIndex=1:ncol(unidata), colWidth=20)
a <- 1
for(e in 1:length(self$ngrams)){
name1 <- as.data.frame(paste("Ngrams of",names(self$ngrams)[[e]]))
addDataFrame(name1, sheet=ngramsheet,startRow= a+3 ,startColumn=2, row.names=FALSE,col.names = FALSE)
addDataFrame(self$ngrams[[e]]$TwoGram, sheet=ngramsheet,startRow= a+4,startColumn= 2, row.names=FALSE,colnamesStyle=csTableColNames)
addDataFrame(self$ngrams[[e]]$ThreeGram, sheet=ngramsheet,startRow= a+4,startColumn=6, row.names=FALSE,colnamesStyle=csTableColNames)
addDataFrame(self$ngrams[[e]]$FourGram, sheet=ngramsheet,startRow= a+4,startColumn=10, row.names=FALSE,colnamesStyle=csTableColNames)
addDataFrame(self$ngrams[[e]]$FiveGram, sheet=ngramsheet,startRow= a+4,startColumn=14, row.names=FALSE,colnamesStyle=csTableColNames)
a <- a + 4
}
saveWorkbook(wb1, savepath)
}
else{
wb1<-loadWorkbook(savepath)
sheet = createSheet(wb1,sheet)
ngramsheet = createSheet(wb1,"Ngrams")
csTableColNames <- CellStyle(wb1) + Font(wb1, isBold=TRUE) + Alignment(wrapText=TRUE, h="ALIGN_CENTER") + Border(color="black", position=c("TOP", "BOTTOM"), pen=c("BORDER_THIN", "BORDER_THICK"))
addDataFrame(unidata, sheet=sheet,startRow=2,startColumn=1, row.names=FALSE,colnamesStyle=csTableColNames)
setColumnWidth(sheet, colIndex=1:ncol(unidata), colWidth=20)
a <- 1
for(e in 1:length(self$ngrams)){
name1 <- as.data.frame(paste("Ngrams of",names(self$ngrams)[[e]]))
addDataFrame(name1, sheet=ngramsheet,startRow= a+3 ,startColumn=2, row.names=FALSE,col.names = FALSE)
addDataFrame(self$ngrams[[e]]$TwoGram, sheet=ngramsheet,startRow= a+4,startColumn= 2, row.names=FALSE,colnamesStyle=csTableColNames)
addDataFrame(self$ngrams[[e]]$ThreeGram, sheet=ngramsheet,startRow= a+4,startColumn=6, row.names=FALSE,colnamesStyle=csTableColNames)
addDataFrame(self$ngrams[[e]]$FourGram, sheet=ngramsheet,startRow= a+4,startColumn=10, row.names=FALSE,colnamesStyle=csTableColNames)
addDataFrame(self$ngrams[[e]]$FiveGram, sheet=ngramsheet,startRow= a+4,startColumn=14, row.names=FALSE,colnamesStyle=csTableColNames)
a <- a + 4
}
saveWorkbook(wb1,savepath)
}
},
saveplot = function(savepath ,breaks = NULL){
require(xlsx)
require(tm)
require(SnowballC)
require(wordcloud)
require(RColorBrewer)
bar_one <- function(column){
df1 <- as.data.frame(table(column))
df <- as.data.frame(sort(table(column),decreasing = T))
df$Rank <- seq(length(df$Freq))
df$Colour[df$Rank <= 3] <- "blue"
df$Colour[df$Rank > 3] <- "black"
df2 <- merge(df,df1,by.x = "column",by.y = "column")
return(df2)
}
rankfreq <- function(column){
df <- as.data.frame(sort(table(column),decreasing = T))
df$Rank <- seq(length(df$Freq))
df$Colour[df$Rank <= 10] <- "blue"
df$Colour[df$Rank > 10] <- "black"
plot(df$Rank,df$Freq,log='xy',type='b',col = df$Colour,xlab = "Rank",ylab = "Frequency",xaxt = 'n',cex.lab=0.75)
axis(1, at=1:length(levels(df$column)),labels=levels(df$column),las = 2)
}
k <- 1
l <- 1
q <- 1
metadata1 <- NULL
metadata <- self$metadata
data <- metadata$data
for(i in 1:length(metadata$columns)){
metadata1 <- rbind(metadata1,as.data.frame(metadata$columns[[i]]))
}
num_var <- as.character(metadata1[which(metadata1$type %in% c("discrete","continuous")),"column_name"])
ord_var <- as.character(metadata1[which(tolower(metadata1$type) == "ordered"),"column_name"])
cat_var <- as.character(metadata1[which(tolower(metadata1$type) == "character"),"column_name"])
text_var <- as.character(metadata1[which(metadata1$type %in% c("text")),"column_name"])
if(file.exists(savepath)){
wb1<-loadWorkbook(savepath)
if(!identical(num_var, character(0))){
sheet_hist = createSheet(wb1,"Univariate-Histogram")
sheet_box = createSheet(wb1,"Univariate-Box Plots")
}
if(!identical(cat_var, character(0))){
sheet_rank = createSheet(wb1,"Univariate-Rank Frequency")
}
if(!identical(ord_var, character(0))){
sheet_bar = createSheet(wb1,"Univariate-Bar Plot")
}
if(!identical(text_var, character(0))){
sheet_wc = createSheet(wb1,"Word cloud")
}
for(i in 1:ncol(data)){
if(names(data)[i] %in% num_var & all(!is.na(data[,i])) ){
q1 <-quantile(data[,i],0.25,na.rm = T)
q3 <- quantile(data[,i],0.75,na.rm =T )
iqr <- q3 - q1
a2 <- q3 + 1.5*iqr
a3 <- q1 - 1.5*iqr
png("boxplot.png", height=1200, width=2000, res=250, pointsize=8)
boxplot(data[,i],col = "blue",main = paste("Boxplot of ",names(data)[i]),xlab = names(data)[i],yaxt = 'n')
axis(side= 2, at= c(min(data[,i]),a3,q1,median(data[,i]),q3,a2,max(data[,i])), labels= c(min(data[,i]),a3,q1,median(data[,i]),q3,a2,max(data[,i])),las = 2)
dev.off()
addPicture("boxplot.png", sheet_box, scale = 1, startRow = q,startColumn = 5)
res<-file.remove("boxplot.png")
if(is.null(breaks)){
breaks <- nclass.FD(data[,i])
}
png("histogram.png", height=1200, width=2000, res=250, pointsize=8)
hist(data[,i],breaks = breaks,col = "blue",main = paste("Histogram of ",names(data)[i]),xlab = names(data)[i])
dev.off()
addPicture("histogram.png", sheet_hist, scale = 1, startRow = q,startColumn = 5)
res<-file.remove("histogram.png")
q <- q + 27
}
else if(names(data)[i] %in% cat_var & length(unique(data[,i])) != 1 & all(!is.na(data[,i]))){
png("rankfreq.png", height=1200, width=2000, res=250, pointsize=8)
rankfreq(data[,i])
title(paste("Rank Frequency plot of",names(data)[i]))
dev.off()
addPicture("rankfreq.png", sheet_rank, scale = 1, startRow = k,startColumn = 5)
res<-file.remove("rankfreq.png")
k <- k + 27
}
else if(names(data)[i] %in% ord_var & length(unique(data[,i])) != 1 & all(!is.na(data[,i]))){
png("barplot.png", height=1200, width=2000, res=250, pointsize=8)
df <- bar_one(data[,i])
barplot(table(data[,i]),col=df$Colour,main = paste("Barplot of ",names(data)[i]),xlab = names(data)[i])
dev.off()
addPicture("barplot.png", sheet_bar, scale = 1, startRow = l,startColumn = 5)
res<-file.remove("barplot.png")
l <- l + 27
}
else if(names(data)[i] %in% text_var & all(!is.na(data[,i]))){
png("wordcloud.png", height=1200, width=2000, res=250, pointsize=8)
corpus <- Corpus(VectorSource(as.character(data[,i])))
corpus_data<-tm_map(corpus,stripWhitespace)
corpus_data<-tm_map(corpus_data,tolower)
corpus_data<-tm_map(corpus_data,removeNumbers)
corpus_data<-tm_map(corpus_data,removePunctuation)
corpus_data<-tm_map(corpus_data,removeWords, stopwords("english"))
corpus_data<-tm_map(corpus_data,removeWords, c("and","the","our","that","for","are","also","more","has","must","have","should","this","with"))
layout(matrix(c(1, 2), nrow=2), heights=c(1, 4))
par(mar=rep(0, 4))
plot.new()
text(x=0.5, y=0.5, paste("Word cloud of",names(data)[i]))
wordcloud(corpus_data, scale=c(4,0.5), min.freq= 5,max.words = 50, random.order=FALSE, rot.per=0.35, use.r.layout=FALSE, colors=brewer.pal(8, "Dark2"))
dev.off()
addPicture("wordcloud.png", sheet_wc, scale = 1, startRow = l,startColumn = 5)
res<-file.remove("wordcloud.png")
l <- l + 27
}
}
saveWorkbook(wb1, savepath)
}
else{
wb1<-createWorkbook(type="xlsx")
if(!identical(num_var, character(0))){
sheet_hist = createSheet(wb1,"Univariate-Histogram")
sheet_box = createSheet(wb1,"Box Plots")
}
if(!identical(cat_var, character(0))){
sheet_rank = createSheet(wb1,"Univariate-Rank Frequency")
}
if(!identical(ord_var, character(0))){
sheet_bar = createSheet(wb1,"Univariate-Bar")
}
if(!identical(text_var, character(0))){
sheet_wc = createSheet(wb1,"Word cloud")
}
for(i in 1:ncol(data)){
if(names(data)[i] %in% num_var & all(!is.na(data[,i]))){
q1 <-quantile(data[,i],0.25,na.rm = T)
q3 <- quantile(data[,i],0.75,na.rm =T )
iqr <- q3 - q1
a2 <- q3 + 1.5*iqr
a3 <- q1 - 1.5*iqr
png("boxplot.png", height=1200, width=2000, res=250, pointsize=8)
boxplot(data[,i],col = "blue",main = paste("Boxplot of ",names(data)[i]),xlab = names(data)[i],yaxt = 'n')
axis(side= 2, at= c(min(data[,i]),a3,q1,median(data[,i]),q3,a2,max(data[,i])), labels= c(min(data[,i]),a3,q1,median(data[,i]),q3,a2,max(data[,i])),las = 2)
dev.off()
addPicture("boxplot.png", sheet_box, scale = 1, startRow = q,startColumn = 5)
res<-file.remove("boxplot.png")
if(is.null(breaks)){
breaks <- nclass.FD(data[,i])
}
png("histogram.png", height=1200, width=2000, res=250, pointsize=8)
hist(data[,i],breaks = breaks,col = "blue",main = paste("Histogram of ",names(data)[i]),xlab = names(data)[i])
dev.off()
addPicture("histogram.png", sheet_hist, scale = 1, startRow = q,startColumn = 5)
res<-file.remove("histogram.png")
q <- q + 27
}
else if(names(data)[i] %in% cat_var & length(unique(data[,i])) != 1 & all(!is.na(data[,i]))){
png("rankfreq.png", height=1200, width=2000, res=250, pointsize=8)
rankfreq(data[,i])
title(paste("Rank Frequency plot of",names(data)[i]))
dev.off()
addPicture("rankfreq.png", sheet_rank, scale = 1, startRow = k,startColumn = 5)
res<-file.remove("rankfreq.png")
k <- k + 27
}
else if(names(data)[i] %in% ord_var & length(unique(data[,i])) != 1 & all(!is.na(data[,i]))){
png("barplot.png", height=1200, width=2000, res=250, pointsize=8)
df <- bar_one(data[,i])
barplot(table(data[,i]),col=df$Colour,main = paste("Barplot of ",names(data)[i]),xlab = names(data)[i])
dev.off()
addPicture("barplot.png", sheet_bar, scale = 1, startRow = l,startColumn = 5)
res<-file.remove("barplot.png")
l <- l + 27
}
else if(names(data)[i] %in% text_var & all(!is.na(data[,i]))){
png("wordcloud.png", height=1200, width=2000, res=250, pointsize=8)
corpus <- Corpus(VectorSource(as.character(data[,i])))
corpus_data<-tm_map(corpus,stripWhitespace)
corpus_data<-tm_map(corpus_data,tolower)
corpus_data<-tm_map(corpus_data,removeNumbers)
corpus_data<-tm_map(corpus_data,removePunctuation)
corpus_data<-tm_map(corpus_data,removeWords, stopwords("english"))
corpus_data<-tm_map(corpus_data,removeWords, c("and","the","our","that","for","are","also","more","has","must","have","should","this","with"))
layout(matrix(c(1, 2), nrow=2), heights=c(1, 4))
par(mar=rep(0, 4))
plot.new()
text(x=0.5, y=0.5, paste("Word cloud of",names(data)[i]))
wordcloud(corpus_data, scale=c(4,0.5), min.freq= 5,max.words = 50, random.order=FALSE, rot.per=0.35, use.r.layout=FALSE, colors=brewer.pal(8, "Dark2"))
dev.off()
addPicture("wordcloud.png", sheet_wc, scale = 1, startRow = l,startColumn = 5)
res<-file.remove("wordcloud.png")
l <- l + 27
}
}
saveWorkbook(wb1, savepath)
}
},
output = function(){
unidata <- NULL
for(i in 1:length(self$columns)){
unidata <- rbind(unidata,as.data.frame(self$columns[[i]]))
}
xyz <- list("univariate"=unidata,"ngrams"=self$ngrams)
return(xyz)
}
)
)
nolancardozo13/eda documentation built on May 12, 2019, 8:47 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
library(R6)
univariate <- R6Class(
"Univariate Analysis",
portable = FALSE,
cloneable = FALSE,
public = list(
columns = list(),
metadata = list(),
ngrams = list(),
initialize = function(metadata,k = 3) {
options(scipen = 999)
dist <- function(var){
a<-cumsum(as.numeric(var))
b<- a/sum(as.numeric(var))
c<- length(which(b<=0.8))
d<- c/length(var)
return(d)
}
findNGrams <- function(df, n) {
ngram <- NGramTokenizer(df, Weka_control(min = n, max = n, delimiters = " \\r\\n\\t.,;:\"()?!"))
ngram2 <- data.frame(table(ngram))
ngram3 <- ngram2[order(ngram2$Freq, decreasing = TRUE),]
colnames(ngram3) <- c("Text","Count")
head(ngram3,10)
}
if(!missing(metadata))
{
self$metadata <- as.list.environment(metadata)
metadata <- as.list.environment(metadata)
data <- metadata$data
metadata1 <- NULL
for(i in 1:length(metadata$columns)){
metadata1 <- rbind(metadata1,as.data.frame(metadata$columns[[i]]))
}
num_var <- as.character(metadata1[which(metadata1$type %in% c("discrete","continuous")),"column_name"])
cat_var <- as.character(metadata1[which(metadata1$type %in% c("character","ordered")),"column_name"])
text_var <- as.character(metadata1[which(metadata1$type %in% c("text")),"column_name"])
self$ngrams <- vector("list", length(text_var))
names(self$ngrams) <- text_var
self$columns <- vector("list", length(num_var))
names(self$columns) <- num_var
for(q in 1:length(text_var)){
self$ngrams[[q]] <- vector("list",4)
names(self$ngrams[[q]]) <- c("TwoGram","ThreeGram","FourGram","FiveGram")
}
mydataframe <- c("column_name","upper_outliers_IQR","lower_outliers_IQR",paste("upper_outliers_",k,"sigma",sep = ""),paste("lower_outliers_",k,"sigma",sep = ""),"concentration","priority","performance","notes")
for(i in 1:ncol(data)){
if(names(data)[i] %in% num_var){
self$columns[[i]] <- setNames(data.frame(matrix(ncol = length(mydataframe), nrow = 0)), mydataframe)
a1 <- names(data)[i]
q1 <-quantile(data[,i],0.25,na.rm = T)
q3 <- quantile(data[,i],0.75,na.rm =T )
iqr <- q3 - q1
a2 <- q3 + 1.5*iqr
a3 <- q1 - 1.5*iqr
a4 <- paste(round(((length(which(data[,i] > a2))*100)/(nrow(data))),1),"%")
a5 <- paste(round(((length(which(data[,i] < a3))*100)/(nrow(data))),1),"%")
a6 <- mean(data[,i],na.rm = T) + k*sd(data[,i],na.rm = T)
a7 <- mean(data[,i],na.rm = T) - k*sd(data[,i],na.rm = T)
a8 <- paste(round(((length(which(data[,i] > a6))*100)/(nrow(data))),1),"%")
a9 <- paste(round(((length(which(data[,i] < a7))*100)/(nrow(data))),1),"%")
a13 <- paste(round((dist(data[,i]) *100)),"%")
a10 <- NA
a11 <- NA
a12 <- NA
self$columns[[i]] <- list("a1" = a1,"a4"=a4,"a5"=a5,"a8"=a8,"a9"=a9,"a10" = a13,"a11"= a10,"a12" = a11,"a13" = a12)
names(self$columns[[i]]) <- mydataframe
rm(a1); rm(a2); rm(a3); rm(a4); rm(a5); rm(a6); rm(a7); rm(a8); rm(a9); rm(a10); rm(a11); rm(a12); rm(a13);
}
else if(names(data)[i] %in% text_var){
for(w in 1:length(text_var)){
self$ngrams[[w]]$TwoGram <- findNGrams(data[,i], 2)
self$ngrams[[w]]$ThreeGram <- findNGrams(data[,i], 3)
self$ngrams[[w]]$FourGram <- findNGrams(data[,i], 4)
self$ngrams[[w]]$FiveGram <- findNGrams(data[,i], 5)
}
}
}
}
},
save = function(savepath,sheet= "Univariate"){
require(xlsx)
unidata <- NULL
for(i in 1:length(self$columns)){
unidata <- rbind(unidata,as.data.frame(self$columns[[i]]))
}
names(unidata) <- gsub("_"," ",names(unidata))
if(!file.exists(savepath)){
wb1 = createWorkbook()
sheet = createSheet(wb1,sheet)
ngramsheet = createSheet(wb1,"Ngrams")
csTableColNames <- CellStyle(wb1) + Font(wb1, isBold=TRUE) + Alignment(wrapText=TRUE, h="ALIGN_CENTER") + Border(color="black", position=c("TOP", "BOTTOM"), pen=c("BORDER_THIN", "BORDER_THICK"))
addDataFrame(unidata, sheet=sheet,startRow=2,startColumn=1, row.names=FALSE,colnamesStyle=csTableColNames)
setColumnWidth(sheet, colIndex=1:ncol(unidata), colWidth=20)
a <- 1
for(e in 1:length(self$ngrams)){
name1 <- as.data.frame(paste("Ngrams of",names(self$ngrams)[[e]]))
addDataFrame(name1, sheet=ngramsheet,startRow= a+3 ,startColumn=2, row.names=FALSE,col.names = FALSE)
addDataFrame(self$ngrams[[e]]$TwoGram, sheet=ngramsheet,startRow= a+4,startColumn= 2, row.names=FALSE,colnamesStyle=csTableColNames)
addDataFrame(self$ngrams[[e]]$ThreeGram, sheet=ngramsheet,startRow= a+4,startColumn=6, row.names=FALSE,colnamesStyle=csTableColNames)
addDataFrame(self$ngrams[[e]]$FourGram, sheet=ngramsheet,startRow= a+4,startColumn=10, row.names=FALSE,colnamesStyle=csTableColNames)
addDataFrame(self$ngrams[[e]]$FiveGram, sheet=ngramsheet,startRow= a+4,startColumn=14, row.names=FALSE,colnamesStyle=csTableColNames)
a <- a + 4
}
saveWorkbook(wb1, savepath)
}
else{
wb1<-loadWorkbook(savepath)
sheet = createSheet(wb1,sheet)
ngramsheet = createSheet(wb1,"Ngrams")
csTableColNames <- CellStyle(wb1) + Font(wb1, isBold=TRUE) + Alignment(wrapText=TRUE, h="ALIGN_CENTER") + Border(color="black", position=c("TOP", "BOTTOM"), pen=c("BORDER_THIN", "BORDER_THICK"))
addDataFrame(unidata, sheet=sheet,startRow=2,startColumn=1, row.names=FALSE,colnamesStyle=csTableColNames)
setColumnWidth(sheet, colIndex=1:ncol(unidata), colWidth=20)
a <- 1
for(e in 1:length(self$ngrams)){
name1 <- as.data.frame(paste("Ngrams of",names(self$ngrams)[[e]]))
addDataFrame(name1, sheet=ngramsheet,startRow= a+3 ,startColumn=2, row.names=FALSE,col.names = FALSE)
addDataFrame(self$ngrams[[e]]$TwoGram, sheet=ngramsheet,startRow= a+4,startColumn= 2, row.names=FALSE,colnamesStyle=csTableColNames)
addDataFrame(self$ngrams[[e]]$ThreeGram, sheet=ngramsheet,startRow= a+4,startColumn=6, row.names=FALSE,colnamesStyle=csTableColNames)
addDataFrame(self$ngrams[[e]]$FourGram, sheet=ngramsheet,startRow= a+4,startColumn=10, row.names=FALSE,colnamesStyle=csTableColNames)
addDataFrame(self$ngrams[[e]]$FiveGram, sheet=ngramsheet,startRow= a+4,startColumn=14, row.names=FALSE,colnamesStyle=csTableColNames)
a <- a + 4
}
saveWorkbook(wb1,savepath)
}
},
saveplot = function(savepath ,breaks = NULL){
require(xlsx)
require(tm)
require(SnowballC)
require(wordcloud)
require(RColorBrewer)
bar_one <- function(column){
df1 <- as.data.frame(table(column))
df <- as.data.frame(sort(table(column),decreasing = T))
df$Rank <- seq(length(df$Freq))
df$Colour[df$Rank <= 3] <- "blue"
df$Colour[df$Rank > 3] <- "black"
df2 <- merge(df,df1,by.x = "column",by.y = "column")
return(df2)
}
rankfreq <- function(column){
df <- as.data.frame(sort(table(column),decreasing = T))
df$Rank <- seq(length(df$Freq))
df$Colour[df$Rank <= 10] <- "blue"
df$Colour[df$Rank > 10] <- "black"
plot(df$Rank,df$Freq,log='xy',type='b',col = df$Colour,xlab = "Rank",ylab = "Frequency",xaxt = 'n',cex.lab=0.75)
axis(1, at=1:length(levels(df$column)),labels=levels(df$column),las = 2)
}
k <- 1
l <- 1
q <- 1
metadata1 <- NULL
metadata <- self$metadata
data <- metadata$data
for(i in 1:length(metadata$columns)){
metadata1 <- rbind(metadata1,as.data.frame(metadata$columns[[i]]))
}
num_var <- as.character(metadata1[which(metadata1$type %in% c("discrete","continuous")),"column_name"])
ord_var <- as.character(metadata1[which(tolower(metadata1$type) == "ordered"),"column_name"])
cat_var <- as.character(metadata1[which(tolower(metadata1$type) == "character"),"column_name"])
text_var <- as.character(metadata1[which(metadata1$type %in% c("text")),"column_name"])
if(file.exists(savepath)){
wb1<-loadWorkbook(savepath)
if(!identical(num_var, character(0))){
sheet_hist = createSheet(wb1,"Univariate-Histogram")
sheet_box = createSheet(wb1,"Univariate-Box Plots")
}
if(!identical(cat_var, character(0))){
sheet_rank = createSheet(wb1,"Univariate-Rank Frequency")
}
if(!identical(ord_var, character(0))){
sheet_bar = createSheet(wb1,"Univariate-Bar Plot")
}
if(!identical(text_var, character(0))){
sheet_wc = createSheet(wb1,"Word cloud")
}
for(i in 1:ncol(data)){
if(names(data)[i] %in% num_var & all(!is.na(data[,i])) ){
q1 <-quantile(data[,i],0.25,na.rm = T)
q3 <- quantile(data[,i],0.75,na.rm =T )
iqr <- q3 - q1
a2 <- q3 + 1.5*iqr
a3 <- q1 - 1.5*iqr
png("boxplot.png", height=1200, width=2000, res=250, pointsize=8)
boxplot(data[,i],col = "blue",main = paste("Boxplot of ",names(data)[i]),xlab = names(data)[i],yaxt = 'n')
axis(side= 2, at= c(min(data[,i]),a3,q1,median(data[,i]),q3,a2,max(data[,i])), labels= c(min(data[,i]),a3,q1,median(data[,i]),q3,a2,max(data[,i])),las = 2)
dev.off()
addPicture("boxplot.png", sheet_box, scale = 1, startRow = q,startColumn = 5)
res<-file.remove("boxplot.png")
if(is.null(breaks)){
breaks <- nclass.FD(data[,i])
}
png("histogram.png", height=1200, width=2000, res=250, pointsize=8)
hist(data[,i],breaks = breaks,col = "blue",main = paste("Histogram of ",names(data)[i]),xlab = names(data)[i])
dev.off()
addPicture("histogram.png", sheet_hist, scale = 1, startRow = q,startColumn = 5)
res<-file.remove("histogram.png")
q <- q + 27
}
else if(names(data)[i] %in% cat_var & length(unique(data[,i])) != 1 & all(!is.na(data[,i]))){
png("rankfreq.png", height=1200, width=2000, res=250, pointsize=8)
rankfreq(data[,i])
title(paste("Rank Frequency plot of",names(data)[i]))
dev.off()
addPicture("rankfreq.png", sheet_rank, scale = 1, startRow = k,startColumn = 5)
res<-file.remove("rankfreq.png")
k <- k + 27
}
else if(names(data)[i] %in% ord_var & length(unique(data[,i])) != 1 & all(!is.na(data[,i]))){
png("barplot.png", height=1200, width=2000, res=250, pointsize=8)
df <- bar_one(data[,i])
barplot(table(data[,i]),col=df$Colour,main = paste("Barplot of ",names(data)[i]),xlab = names(data)[i])
dev.off()
addPicture("barplot.png", sheet_bar, scale = 1, startRow = l,startColumn = 5)
res<-file.remove("barplot.png")
l <- l + 27
}
else if(names(data)[i] %in% text_var & all(!is.na(data[,i]))){
png("wordcloud.png", height=1200, width=2000, res=250, pointsize=8)
corpus <- Corpus(VectorSource(as.character(data[,i])))
corpus_data<-tm_map(corpus,stripWhitespace)
corpus_data<-tm_map(corpus_data,tolower)
corpus_data<-tm_map(corpus_data,removeNumbers)
corpus_data<-tm_map(corpus_data,removePunctuation)
corpus_data<-tm_map(corpus_data,removeWords, stopwords("english"))
corpus_data<-tm_map(corpus_data,removeWords, c("and","the","our","that","for","are","also","more","has","must","have","should","this","with"))
layout(matrix(c(1, 2), nrow=2), heights=c(1, 4))
par(mar=rep(0, 4))
plot.new()
text(x=0.5, y=0.5, paste("Word cloud of",names(data)[i]))
wordcloud(corpus_data, scale=c(4,0.5), min.freq= 5,max.words = 50, random.order=FALSE, rot.per=0.35, use.r.layout=FALSE, colors=brewer.pal(8, "Dark2"))
dev.off()
addPicture("wordcloud.png", sheet_wc, scale = 1, startRow = l,startColumn = 5)
res<-file.remove("wordcloud.png")
l <- l + 27
}
}
saveWorkbook(wb1, savepath)
}
else{
wb1<-createWorkbook(type="xlsx")
if(!identical(num_var, character(0))){
sheet_hist = createSheet(wb1,"Univariate-Histogram")
sheet_box = createSheet(wb1,"Box Plots")
}
if(!identical(cat_var, character(0))){
sheet_rank = createSheet(wb1,"Univariate-Rank Frequency")
}
if(!identical(ord_var, character(0))){
sheet_bar = createSheet(wb1,"Univariate-Bar")
}
if(!identical(text_var, character(0))){
sheet_wc = createSheet(wb1,"Word cloud")
}
for(i in 1:ncol(data)){
if(names(data)[i] %in% num_var & all(!is.na(data[,i]))){
q1 <-quantile(data[,i],0.25,na.rm = T)
q3 <- quantile(data[,i],0.75,na.rm =T )
iqr <- q3 - q1
a2 <- q3 + 1.5*iqr
a3 <- q1 - 1.5*iqr
png("boxplot.png", height=1200, width=2000, res=250, pointsize=8)
boxplot(data[,i],col = "blue",main = paste("Boxplot of ",names(data)[i]),xlab = names(data)[i],yaxt = 'n')
axis(side= 2, at= c(min(data[,i]),a3,q1,median(data[,i]),q3,a2,max(data[,i])), labels= c(min(data[,i]),a3,q1,median(data[,i]),q3,a2,max(data[,i])),las = 2)
dev.off()
addPicture("boxplot.png", sheet_box, scale = 1, startRow = q,startColumn = 5)
res<-file.remove("boxplot.png")
if(is.null(breaks)){
breaks <- nclass.FD(data[,i])
}
png("histogram.png", height=1200, width=2000, res=250, pointsize=8)
hist(data[,i],breaks = breaks,col = "blue",main = paste("Histogram of ",names(data)[i]),xlab = names(data)[i])
dev.off()
addPicture("histogram.png", sheet_hist, scale = 1, startRow = q,startColumn = 5)
res<-file.remove("histogram.png")
q <- q + 27
}
else if(names(data)[i] %in% cat_var & length(unique(data[,i])) != 1 & all(!is.na(data[,i]))){
png("rankfreq.png", height=1200, width=2000, res=250, pointsize=8)
rankfreq(data[,i])
title(paste("Rank Frequency plot of",names(data)[i]))
dev.off()
addPicture("rankfreq.png", sheet_rank, scale = 1, startRow = k,startColumn = 5)
res<-file.remove("rankfreq.png")
k <- k + 27
}
else if(names(data)[i] %in% ord_var & length(unique(data[,i])) != 1 & all(!is.na(data[,i]))){
png("barplot.png", height=1200, width=2000, res=250, pointsize=8)
df <- bar_one(data[,i])
barplot(table(data[,i]),col=df$Colour,main = paste("Barplot of ",names(data)[i]),xlab = names(data)[i])
dev.off()
addPicture("barplot.png", sheet_bar, scale = 1, startRow = l,startColumn = 5)
res<-file.remove("barplot.png")
l <- l + 27
}
else if(names(data)[i] %in% text_var & all(!is.na(data[,i]))){
png("wordcloud.png", height=1200, width=2000, res=250, pointsize=8)
corpus <- Corpus(VectorSource(as.character(data[,i])))
corpus_data<-tm_map(corpus,stripWhitespace)
corpus_data<-tm_map(corpus_data,tolower)
corpus_data<-tm_map(corpus_data,removeNumbers)
corpus_data<-tm_map(corpus_data,removePunctuation)
corpus_data<-tm_map(corpus_data,removeWords, stopwords("english"))
corpus_data<-tm_map(corpus_data,removeWords, c("and","the","our","that","for","are","also","more","has","must","have","should","this","with"))
layout(matrix(c(1, 2), nrow=2), heights=c(1, 4))
par(mar=rep(0, 4))
plot.new()
text(x=0.5, y=0.5, paste("Word cloud of",names(data)[i]))
wordcloud(corpus_data, scale=c(4,0.5), min.freq= 5,max.words = 50, random.order=FALSE, rot.per=0.35, use.r.layout=FALSE, colors=brewer.pal(8, "Dark2"))
dev.off()
addPicture("wordcloud.png", sheet_wc, scale = 1, startRow = l,startColumn = 5)
res<-file.remove("wordcloud.png")
l <- l + 27
}
}
saveWorkbook(wb1, savepath)
}
},
output = function(){
unidata <- NULL
for(i in 1:length(self$columns)){
unidata <- rbind(unidata,as.data.frame(self$columns[[i]]))
}
xyz <- list("univariate"=unidata,"ngrams"=self$ngrams)
return(xyz)
}
)
)
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.