Cell Phone Analysis

cellphone analysis.R

library(jsonlite)

cp = fromJSON(txt = "Cell Phone Data.txt", simplifyDataFrame = TRUE)

num.atts = c(4,9,11,12,13,14,15,16,18,22)

cp[,num.atts] = sapply(cp[,num.atts], function (x) as.numeric(x))
cp$aspect.ratio = cp$att_pixels_y / cp$att_pixels_x
cp$isSmartPhone = ifelse(grepl("smart|iphone|blackberry", cp$name, ignore.case=TRUE) == TRUE | cp$att_screen_size >= 4, "Yes", "No")

library(ggplot2)

library(ggthemr)

library(scales)
ggthemr("camoflauge")

ggplot(cp, aes(x=att_brand, y=price)) + geom_boxplot() + ggtitle("Mobile Phone Price by Brand") + theme(axis.text.x=element_text(angle=90, size=14, vjust=0.5), axis.text.y=element_text(size=14), axis.title.x=element_text(size=15), axis.title.y=element_text(size=15), plot.title=element_text(size=17)) + scale_y_continuous(labels=dollar, name="Price (USD?)") + scale_x_discrete("Brand") 

ggplot(cp, aes(x=att_weight, y=price)) + geom_point(size=3) + ggtitle("Mobile Phone Price by Weight") + theme(axis.text.x=element_text(size=14, vjust=0.5), axis.text.y=element_text(size=14), axis.title.x=element_text(size=15), axis.title.y=element_text(size=15), plot.title=element_text(size=17)) + scale_y_continuous(labels=dollar, name="Price (USD?)") + scale_x_continuous("Weight (oz)") + stat_smooth(se=FALSE)

ggplot(cp, aes(x=att_screen_size, y=price)) + geom_point(size=3) + ggtitle("Mobile Phone Price by Screen Size") + theme(axis.text.x=element_text(size=14, vjust=0.5), axis.text.y=element_text(size=14), axis.title.x=element_text(size=15), axis.title.y=element_text(size=15), plot.title=element_text(size=17)) + scale_y_continuous(labels=dollar, name="Price (USD?)") + scale_x_continuous("Screen Size (in)") + stat_smooth(se=FALSE)

ggplot(cp, aes(x=att_ram, y=price)) + geom_point(size=3) + ggtitle("Mobile Phone Price by Amount of RAM") + theme(axis.text.x=element_text(size=14, vjust=0.5), axis.text.y=element_text(size=14), axis.title.x=element_text(size=15), axis.title.y=element_text(size=15), plot.title=element_text(size=17)) + scale_y_continuous(labels=dollar, name="Price (USD?)") + scale_x_continuous("RAM (gb)") + stat_smooth(se=FALSE)

ggplot(cp, aes(x=att_sd_card, y=price)) + geom_point(size=3) + ggtitle("Mobile Phone Price by SD Card Capacity") + theme(axis.text.x=element_text(size=14, vjust=0.5), axis.text.y=element_text(size=14), axis.title.x=element_text(size=15), axis.title.y=element_text(size=15), plot.title=element_text(size=17)) + scale_y_continuous(labels=dollar, name="Price (USD?)") + scale_x_continuous("SD Card Capacity (gb)") + stat_smooth(se=FALSE)

ggplot(cp, aes(x=ifelse(cp$att_dual_sim == 1, "Yes", "No"), y=price)) + geom_boxplot() + ggtitle("Mobile Phone Price by Dual Sim") + theme(axis.text.x=element_text(size=14, vjust=0.5), axis.text.y=element_text(size=14), axis.title.x=element_text(size=15), axis.title.y=element_text(size=15), plot.title=element_text(size=17)) + scale_y_continuous(labels=dollar, name="Price (USD?)") + scale_x_discrete("Has Dual Sim Card?") 

ggplot(cp, aes(x=att_storage, y=price)) + geom_point(size=3) + ggtitle("Mobile Phone Price by Storage Capacity") + theme(axis.text.x=element_text(size=14, vjust=0.5), axis.text.y=element_text(size=14), axis.title.x=element_text(size=15), axis.title.y=element_text(size=15), plot.title=element_text(size=17)) + scale_y_continuous(labels=dollar, name="Price (USD?)") + scale_x_continuous("Storage Capacity (gb)") + stat_smooth(se=FALSE)

ggplot(cp, aes(x=att_battery_mah, y=price)) + geom_point(size=3) + ggtitle("Mobile Phone Price by Battery Capacity") + theme(axis.text.x=element_text(size=14, vjust=0.5), axis.text.y=element_text(size=14), axis.title.x=element_text(size=15), axis.title.y=element_text(size=15), plot.title=element_text(size=17)) + scale_y_continuous(labels=dollar, name="Price (USD?)") + scale_x_continuous("Battery Capacity (mAh)") + stat_smooth(se=FALSE)

ggplot(cp, aes(x=aspect.ratio, y=price)) + geom_point(size=3) + ggtitle("Mobile Phone Price by Aspect Ratio") + theme(axis.text.x=element_text(size=14, vjust=0.5), axis.text.y=element_text(size=14), axis.title.x=element_text(size=15), axis.title.y=element_text(size=15), plot.title=element_text(size=17)) + scale_y_continuous(labels=dollar, name="Price (USD?)") + scale_x_continuous("Aspect Ratio (Y Pixels / X Pixels)") + stat_smooth(se=FALSE)

ggplot(cp, aes(x=isSmartPhone, y=price)) + geom_boxplot() + ggtitle("Mobile Phone Price by Smart Phone Status") + theme(axis.text.x=element_text(size=14, vjust=0.5), axis.text.y=element_text(size=14), axis.title.x=element_text(size=15), axis.title.y=element_text(size=15), plot.title=element_text(size=17)) + scale_y_continuous(labels=dollar, name="Price (USD?)") + scale_x_discrete("Is it a Smart Phone?") 

ggplot(cp, aes(x=att_os, y=price)) + geom_boxplot() + ggtitle("Mobile Phone Price by Operating System") + theme(axis.text.x=element_text(size=14, vjust=0.5), axis.text.y=element_text(size=14), axis.title.x=element_text(size=15), axis.title.y=element_text(size=15), plot.title=element_text(size=17)) + scale_y_continuous(labels=dollar, name="Price (USD?)") + scale_x_discrete("Operating System") 

library(caret)

control = trainControl(method="cv")
in_train = createDataPartition(cp$price, p=.8, list=FALSE)
model.gbm = train(price ~ att_brand + att_weight + att_screen_size +
                    att_ram + att_sd_card + att_dual_sim +
                    att_storage + att_battery_mah + att_os, data=cp,
                  method="gbm", trControl=control, verbose=FALSE, subset=in_train)
cp$att_brand = factor(cp$)
cp.test = cp[-in_train,]
cp.test = subset(cp.test, att_brand != 'TOTO')
cp.test = na.omit(cp.test)
cp.test$pred.price = predict(model.gbm, cp.test)

ggplot(cp.test, aes(x=pred.price, y=price)) + geom_point(size=3) + ggtitle("Mobile Phone Price by Predicted Price") + theme(axis.text.x=element_text(size=14, vjust=0.5), axis.text.y=element_text(size=14), axis.title.x=element_text(size=15), axis.title.y=element_text(size=15), plot.title=element_text(size=17)) + scale_y_continuous(labels=dollar, name="Price (USD?)") + scale_x_continuous("Predicted Price", labels=dollar) + geom_abline(intercept=0, slope=1, colour="yellow") + stat_smooth(se=FALSE)

Hello inkhorn, I can't download the file "Cell Phone Data.txt". Can you please share the link? If not, can you please teach how to get the data from the url? Appreciate the guidance. Thanks.