Pratik Bhujade

01/11/2019

Question 1

1. Created the dataset, and cleaning it

# Importing the tidyR and zoo Library for analysis

library(tidyr)
library(zoo)

##
## Attaching package: zoo

## The following objects are masked from package:base :

##
## as.Date, as.Date.numeric

# Inserting the dataset

df <- read.csv( DiamondData.csv )
# Removing the spelling mistake in cuts.
df_cleaning <- df
df_cleaning$cut[df_cleaning$cut == "Very Geod"] <-"Very Good"
# Insert means rows in rows with na value
df_clean <- data.frame( cut = df_cleaning$cut, color = df_cleaning$color, clarity = df_cleaning$clar
df_clean$carat <- na.aggregate(df_cleaning$carat)
df_clean$table <- na.aggregate(df_cleaning$table)
df_clean$x <- na.aggregate(df_cleaning$x)
df_clean$y <- na.aggregate(df_cleaning$y)
df_clean$z <- na.aggregate(df_cleaning$z)
df_cleaning$depthh <- (2*df_clean$z/(df_clean$x+df_clean$y))*100
df_clean$depth <- na.aggregate(df_cleaning$depthh)
df_clean$price <- na.aggregate(df_cleaning$price)
df_cleaned <- df_clean
#names(df_cleaned) <- c( cut , color , clarity , carat , table , x , y , z , depth )
summary(df_clean)

## cut color clarity carat

## Fair : 1480 D: 6264 SI1 :12120 Min. : 0.200
## Good : 4559 E: 9066 VS2 :11406 1st Qu.: 0.400
## Ideal :19918 F: 8837 SI2 : 8486 Median : 0.700
## Premium :12826 G:10493 VS1 : 7563 Mean : 0.907
## Very Geod: 0 H: 7705 VVS2 : 4692 3rd Qu.: 1.050
## Very Good:11217 I: 5028 VVS1 : 3377 Max. :49.990
## J: 2607 (Other): 2356
## table x y z
## Min. :43.00 Min. : 0.000 Min. : 0.000 Min. : 0.00
## 1st Qu.:56.00 1st Qu.: 4.710 1st Qu.: 4.720 1st Qu.: 2.91

1
## Median :57.00 Median : 5.700 Median : 5.720 Median : 3.53
## Mean :57.46 Mean : 5.732 Mean : 5.734 Mean : 3.54
## 3rd Qu.:59.00 3rd Qu.: 6.540 3rd Qu.: 6.530 3rd Qu.: 4.03
## Max. :95.00 Max. :10.230 Max. :31.800 Max. :31.80
##
## depth price
## Min. : 0.00 Min. : 326
## 1st Qu.: 61.03 1st Qu.: 954
## Median : 61.84 Median : 2426
## Mean : 61.77 Mean : 3939
## 3rd Qu.: 62.54 3rd Qu.: 5315
## Max. :619.28 Max. :18823
##

Question 2

Summary and Plots

## cut color clarity carat

## Fair : 1480 D: 6264 SI1 :12120 Min. : 0.200
## Good : 4559 E: 9066 VS2 :11406 1st Qu.: 0.400
## Ideal :19918 F: 8837 SI2 : 8486 Median : 0.700
## Premium :12826 G:10493 VS1 : 7563 Mean : 0.907
## Very Geod: 0 H: 7705 VVS2 : 4692 3rd Qu.: 1.050
## Very Good:11217 I: 5028 VVS1 : 3377 Max. :49.990
## J: 2607 (Other): 2356
## table x y z
## Min. :43.00 Min. : 0.000 Min. : 0.000 Min. : 0.00
## 1st Qu.:56.00 1st Qu.: 4.710 1st Qu.: 4.720 1st Qu.: 2.91
## Median :57.00 Median : 5.700 Median : 5.720 Median : 3.53
## Mean :57.46 Mean : 5.732 Mean : 5.734 Mean : 3.54
## 3rd Qu.:59.00 3rd Qu.: 6.540 3rd Qu.: 6.530 3rd Qu.: 4.03
## Max. :95.00 Max. :10.230 Max. :31.800 Max. :31.80
##
## depth price
## Min. : 0.00 Min. : 326
## 1st Qu.: 61.03 1st Qu.: 954
## Median : 61.84 Median : 2426
## Mean : 61.77 Mean : 3939
## 3rd Qu.: 62.54 3rd Qu.: 5315
## Max. :619.28 Max. :18823
##

2
 Cut − Pie Chart

Ideal

Good

Fair

Very Good
Premium
Very Geod

3
 Color − Bar Chart
10000
Frequency

6000
2000
0

D E F G H I J

Colors

4
 Histogram − Depth
50000
30000
Frequency

10000
0

0 100 200 300 400 500 600

Depths

5
 Scatterplot − Carats vs Depth
500
Depth

300
100
0

0 10 20 30 40 50

Carats

## Question 3 ### Part A Histogram of price along with the summary Statistics

# importing psych library for statistics

library(psych)
# histogram for prices
hist(df_cleaned$price, xlab = "Price", ylab = "Distribution", main = "Histogram of Price")

6
 Histogram of Price
12000
8000
Distribution

4000
0

0 5000 10000 15000

Price

# Detailed summary statistics from psych library

describe(df_cleaned$price)

## vars n mean sd median trimmed mad min max range skew

## X1 1 50000 3938.8 3984.52 2426 3166.3 2501.15 326 18823 18497 1.62
## kurtosis se
## X1 2.18 17.82

The Histogram indicates that the distribution is positively skewed, and concentrated to the left. It also
indicates that the price is not a normal distribtion, and the higest number of units lie in the distribution
that is less than 5000 price. The summary statistics indicate that mean is 3938.64, and the media is 2401.
Moreover the standard deviation is 3994.8.

Part B

# Low range 0-2401, first quartile

price_low <- 2401
# Medium Range from 2402-5345, Third quartile
price_med <- 5345
# High range greater than 5345
low <- subset(df_cleaned, price <= price_low)
med <- subset(df_cleaned, price <= price_med)
high <- subset(df_cleaned, price > price_med)

7
# Summary statistics for low, medium and high
summary(low)

## cut color clarity carat

## Fair : 512 D:3595 VS2 :6005 Min. : 0.200
## Good : 1979 E:5309 SI1 :5335 1st Qu.: 0.320
## Ideal :11636 F:4532 VS1 :4129 Median : 0.400
## Premium : 5538 G:5442 VVS2 :3057 Mean : 0.537
## Very Geod: 0 H:3214 VVS1 :2520 3rd Qu.: 0.520
## Very Good: 5211 I:1905 SI2 :2352 Max. :49.728
## J: 879 (Other):1478
## table x y z
## Min. :44.00 Min. :0.000 Min. : 0.00 Min. : 0.000
## 1st Qu.:56.00 1st Qu.:4.390 1st Qu.: 4.40 1st Qu.: 2.710
## Median :57.00 Median :4.710 Median : 4.73 Median : 2.910
## Mean :57.11 Mean :4.801 Mean : 4.81 Mean : 2.969
## 3rd Qu.:58.00 3rd Qu.:5.180 3rd Qu.: 5.18 3rd Qu.: 3.200
## Max. :79.00 Max. :6.860 Max. :31.80 Max. :31.800
##
## depth price
## Min. : 0.00 Min. : 326
## 1st Qu.: 61.07 1st Qu.: 694
## Median : 61.79 Median : 949
## Mean : 61.80 Mean :1126
## 3rd Qu.: 62.42 3rd Qu.:1574
## Max. :619.28 Max. :2401
##

summary(med)

## cut color clarity carat

## Fair : 1118 D:5207 SI1 :9238 Min. : 0.2000
## Good : 3509 E:7625 VS2 :8110 1st Qu.: 0.3400
## Ideal :15691 F:6902 SI2 :6157 Median : 0.5200
## Premium : 8840 G:7542 VS1 :5468 Mean : 0.6955
## Very Geod: 0 H:5397 VVS2 :3700 3rd Qu.: 0.7600
## Very Good: 8420 I:3274 VVS1 :2961 Max. :49.9904
## J:1631 (Other):1944
## table x y z
## Min. :43.00 Min. :0.000 Min. : 0.000 Min. : 0.000
## 1st Qu.:56.00 1st Qu.:4.510 1st Qu.: 4.520 1st Qu.: 2.780
## Median :57.00 Median :5.170 Median : 5.180 Median : 3.200
## Mean :57.36 Mean :5.262 Mean : 5.268 Mean : 3.253
## 3rd Qu.:59.00 3rd Qu.:5.870 3rd Qu.: 5.870 3rd Qu.: 3.620
## Max. :79.00 Max. :8.540 Max. :31.800 Max. :31.800
##
## depth price
## Min. : 0.00 Min. : 326
## 1st Qu.: 61.05 1st Qu.: 812
## Median : 61.83 Median :1588
## Mean : 61.80 Mean :2030
## 3rd Qu.: 62.53 3rd Qu.:3053
## Max. :619.28 Max. :5345

8
##

summary(high)

## cut color clarity carat table

## Fair : 362 D:1057 VS2 :3296 Min. : 0.630 Min. :50.00
## Good :1050 E:1441 SI1 :2882 1st Qu.: 1.090 1st Qu.:56.00
## Ideal :4227 F:1935 SI2 :2329 Median : 1.330 Median :58.00
## Premium :3986 G:2951 VS1 :2095 Mean : 1.547 Mean :57.79
## Very Geod: 0 H:2308 VVS2 : 992 3rd Qu.: 1.600 3rd Qu.:59.00
## Very Good:2797 I:1754 VVS1 : 416 Max. :49.521 Max. :95.00
## J: 976 (Other): 412
## x y z depth
## Min. : 0.000 Min. : 0.000 Min. :0.000 Min. : 0.00
## 1st Qu.: 6.630 1st Qu.: 6.630 1st Qu.:4.080 1st Qu.:60.95
## Median : 7.050 Median : 7.050 Median :4.340 Median :61.88
## Mean : 7.154 Mean : 7.143 Mean :4.407 Mean :61.68
## 3rd Qu.: 7.530 3rd Qu.: 7.520 3rd Qu.:4.630 3rd Qu.:62.58
## Max. :10.230 Max. :10.160 Max. :6.720 Max. :80.57
##
## price
## Min. : 5346
## 1st Qu.: 6609
## Median : 8696
## Mean : 9713
## 3rd Qu.:12183
## Max. :18823
##

Part C

Box plot for different cuts

# boxplot for prices, based on different cuts

boxplot(df_cleaned$price~df_cleaned$cut, xlab = "Cuts", ylab = "Prices",main = "Price distribution of cu

9
 Price distribution of cuts
15000
10000
Prices

5000
0

Fair Good Ideal Premium Very Good

Cuts

Part D

# creating the dataframe for numerical variables

n_df <- data.frame(df_cleaned$carat,df_cleaned$price,df_cleaned$depth,df_cleaned$table,df_cleaned$x,df_c
# Correlation for numerical variables
cor(n_df)

## df_cleaned.carat df_cleaned.price df_cleaned.depth

## df_cleaned.carat 1.000000000 0.2132887 -0.009724909
## df_cleaned.price 0.213288669 1.0000000 -0.020770698
## df_cleaned.depth -0.009724909 -0.0207707 1.000000000
## df_cleaned.table 0.039761303 0.1270487 -0.132510964
## df_cleaned.x 0.223011294 0.8801776 -0.038237138
## df_cleaned.y 0.221418092 0.8756857 -0.050007328
## df_cleaned.z 0.214743207 0.8541778 0.220431188
## df_cleaned.table df_cleaned.x df_cleaned.y df_cleaned.z
## df_cleaned.carat 0.0397613 0.22301129 0.22141809 0.2147432
## df_cleaned.price 0.1270487 0.88017761 0.87568573 0.8541778
## df_cleaned.depth -0.1325110 -0.03823714 -0.05000733 0.2204312
## df_cleaned.table 1.0000000 0.19389482 0.18654068 0.1495847
## df_cleaned.x 0.1938948 1.00000000 0.98714681 0.9629605
## df_cleaned.y 0.1865407 0.98714681 1.00000000 0.9570021
## df_cleaned.z 0.1495847 0.96296045 0.95700213 1.0000000

10
Question 4

Frequencies and Scatterplots

# Importing ggplot for visualization

library(ggplot2)

##
## Attaching package: ggplot2

## The following objects are masked from package:psych :

##
## %+%, alpha

# SUmmary for different cuts

summary(df_cleaned$cut)

## Fair Good Ideal Premium Very Geod Very Good

## 1480 4559 19918 12826 0 11217

# Summary for different clarities

summary(df_cleaned$clarity)

## I1 IF SI1 SI2 VS1 VS2 VVS1 VVS2

## 690 1666 12120 8486 7563 11406 3377 4692

# Relationship between carat and price on the basis of cuts

ggplot(df_cleaned, aes(price, carat, color = cut)) + geom_point()

11
 50

40

cut
30 Fair
Good
carat

Ideal

20 Premium
Very Good

10

0 5000 10000 15000

price

# Relationship between carat and price on the basis of clarities

ggplot(df_cleaned, aes(price, carat, color = clarity)) + geom_point()

12
 50

40
clarity
I1
IF
30
SI1
carat

SI2
VS1
20
VS2
VVS1
VVS2
10

0 5000 10000 15000

price
## Question 5 ### Part A

# creating volume dataframe

volume =data.frame(df_cleaned$x, df_cleaned$y, df_cleaned$z)
# putting names in it
names(volume) <- c( x , y , z )
# calculating volume
volume$volume <- volume$x*volume$y*volume$z
# adding price to the mix
volume$price <- df_cleaned$price
# ploting volume and price
ggplot(volume, aes(volume, price)) + geom_point(color= blue ) + geom_smooth(method = lm , formula = y ~

13
 40000

30000
price

20000

10000

0 200 400 600 800

volume
This indicates a polynomial regression, and it indicates that the regression is increasing in magnitude along
with being positive.

Part B

Relationship between carat and volume

# adding carats to the mix

volume$carat <- df_cleaned$carat
# plotting volume and carat
ggplot(volume, aes(volume, carat)) + geom_point(color= red ) + geom_smooth(method = lm , formula = y ~

14
 50

40

30
carat

20

10

0 200 400 600 800

volume
No linear relationship between volum and carats.

Part C

Relationship between table and depth

# Adding table and depth to the mix

volume$table <- df_cleaned$table
volume$depth <- df_cleaned$depth
# ploting table and depth
ggplot(volume, aes(depth, table)) + geom_point(color= green ) + geom_smooth(method = lm , formula = y ~

15
 75
table

50

25

0 200 400 600

depth
Negative relationship between depth and table ### Part D Relationship of table with all the variables

# finding the correlation between all the added variables

cor(volume)

## x y z volume price
## x 1.00000000 0.98714681 0.9629605 0.97326375 0.8801776
## y 0.98714681 1.00000000 0.9570021 0.97176259 0.8756857
## z 0.96296045 0.95700213 1.0000000 0.96191265 0.8541778
## volume 0.97326375 0.97176259 0.9619127 1.00000000 0.9162446
## price 0.88017761 0.87568573 0.8541778 0.91624455 1.0000000
## carat 0.22301129 0.22141809 0.2147432 0.22671334 0.2132887
## table 0.19389482 0.18654068 0.1495847 0.17033926 0.1270487
## depth -0.03823714 -0.05000733 0.2204312 0.03456058 -0.0207707
## carat table depth
## x 0.223011294 0.1938948 -0.038237138
## y 0.221418092 0.1865407 -0.050007328
## z 0.214743207 0.1495847 0.220431188
## volume 0.226713342 0.1703393 0.034560579
## price 0.213288669 0.1270487 -0.020770698
## carat 1.000000000 0.0397613 -0.009724909
## table 0.039761303 1.0000000 -0.132510964
## depth -0.009724909 -0.1325110 1.000000000

# ploting volume, price and carats with table

ggplot(volume, aes(volume, table)) + geom_point(color= pink ) + geom_smooth(method = lm , formula = y ~

16
 90

80
table

70

60

50

0 200 400 600 800

volume

ggplot(volume, aes(price, table)) + geom_point(color= orange ) + geom_smooth(method = lm , formula = y

17
 90

80
table

70

60

50

0 5000 10000 15000

price

ggplot(volume, aes(carat, table)) + geom_point(color= purple ) + geom_smooth(method = lm , formula = y

18
 90

80
table

70

60

50

0 10 20 30 40 50
carat

19