Data Visualization

Part II

Agenda

We are going to visually analyze two datasets and see if we can tell stories from the visuals.

Setting up

Let’s first load the ggplot2 package:

if (!require(ggplot2)) {
  install.packages("ggplot2") # install if not already installed
}
library (ggplot2)

Titanic Survival

Load and explore the dataset

titanic = read.csv ("https://raw.githubusercontent.com/ahmedmoustafa/datasets/main/titanic/titanic.csv")
head(titanic)

name	survived	sex	age	class
Allen, Miss. Elisabeth Walton	yes	female	29.0000	1st
Allison, Master. Hudson Trevor	yes	male	0.9167	1st
Allison, Miss. Helen Loraine	no	female	2.0000	1st
Allison, Mr. Hudson Joshua Crei	no	male	30.0000	1st
Allison, Mrs. Hudson J C (Bessi	no	female	25.0000	1st
Anderson, Mr. Harry	yes	male	48.0000	1st

titanic$survived = factor(titanic$survived)
titanic$sex = factor(titanic$sex)
titanic$class = factor(titanic$class, levels = c("1st", "2nd", "3rd"))
head(titanic)

name	survived	sex	age	class
Allen, Miss. Elisabeth Walton	yes	female	29.0000	1st
Allison, Master. Hudson Trevor	yes	male	0.9167	1st
Allison, Miss. Helen Loraine	no	female	2.0000	1st
Allison, Mr. Hudson Joshua Crei	no	male	30.0000	1st
Allison, Mrs. Hudson J C (Bessi	no	female	25.0000	1st
Anderson, Mr. Harry	yes	male	48.0000	1st

summary(titanic)

name	survived	sex	age	class
Length:1309	no :809	female:466	Min. : 0.1667	1st:323
Class :character	yes:500	male :843	1st Qu.:21.0000	2nd:277
Mode :character	NA	NA	Median :28.0000	3rd:709
NA	NA	NA	Mean :29.8811	NA
NA	NA	NA	3rd Qu.:39.0000	NA
NA	NA	NA	Max. :80.0000	NA
NA	NA	NA	NA’s :263	NA

Survival by class

ggplot(titanic) +
  geom_bar(aes(x = class, fill = survived), position = "dodge")

ggplot(titanic) +
  geom_bar(aes(x = class, fill = survived), position = "fill")

Survival by sex

ggplot(titanic) +
  geom_bar(aes(x = sex, fill = survived), position = "dodge")

ggplot(titanic) +
  geom_bar(aes(x = sex, fill = survived), position = "fill")

Survival by age

ggplot(titanic) +
  geom_density(aes(x = age, fill = survived), alpha = 0.5)

ggplot(titanic) +
  geom_boxplot(aes(x = survived, y = age))

ggplot(titanic) +
  geom_violin(aes(x = survived, y = age))

Survival by age & sex

ggplot(titanic) +
  geom_boxplot(aes(x = sex, y = age, fill = survived), alpha = 0.5)

Survival by class & age

ggplot(titanic) +
  geom_boxplot(aes(x = class, y = age, fill = survived), alpha = 0.5)

Smoking and Pregnancy

Load and explore the dataset

smoking = read.csv("https://raw.githubusercontent.com/ahmedmoustafa/datasets/main/smoking/smoking.csv")
head(smoking)

id	date	gestation	weight	parity	mom.race	mom.age	mom.edu	mom.height	mom.weight	dad.race	dad.age	dad.edu	dad.height	dad.weight	marital	income	smoke	quit.time	cigs
15	1411	284	120	1	asian	27	5	62	100	asian	31	5	65	110	1	1	never	0	0
20	1499	282	113	2	white	33	5	64	135	white	38	5	70	148	1	4	never	0	0
100	1673	286	136	4	white	25	2	62	93	white	28	2	64	130	1	4	until_pregnancy	2	2
129	1562	245	132	2	black	23	1	65	140	black	23	4	71	192	1	2	never	0	0
142	1408	289	120	3	white	25	4	62	125	white	26	1	70	180	0	2	never	0	0
171	1593	282	144	4	white	32	2	64	124	white	36	1	74	185	1	2	now	1	1

smoking$parity = factor(smoking$parity)
smoking$mom.race = factor(smoking$mom.race)
smoking$mom.edu = factor(smoking$mom.edu)
smoking$dad.race = factor(smoking$dad.race)
smoking$dad.edu = factor(smoking$dad.edu)
smoking$marital = factor(smoking$marital)
smoking$income = factor(smoking$income)
smoking$smoke = factor(smoking$smoke)
smoking$quit.time = factor(smoking$quit.time)
smoking$cigs = factor(smoking$cigs)
head(smoking)

id	date	gestation	weight	parity	mom.race	mom.age	mom.edu	mom.height	mom.weight	dad.race	dad.age	dad.edu	dad.height	dad.weight	marital	income	smoke	quit.time	cigs
15	1411	284	120	1	asian	27	5	62	100	asian	31	5	65	110	1	1	never	0	0
20	1499	282	113	2	white	33	5	64	135	white	38	5	70	148	1	4	never	0	0
100	1673	286	136	4	white	25	2	62	93	white	28	2	64	130	1	4	until_pregnancy	2	2
129	1562	245	132	2	black	23	1	65	140	black	23	4	71	192	1	2	never	0	0
142	1408	289	120	3	white	25	4	62	125	white	26	1	70	180	0	2	never	0	0
171	1593	282	144	4	white	32	2	64	124	white	36	1	74	185	1	2	now	1	1

summary(smoking)

id	date	gestation	weight	parity	mom.race	mom.age	mom.edu	mom.height	mom.weight	dad.race	dad.age	dad.edu	dad.height	dad.weight	marital	income	smoke	quit.time	cigs
Min. : 15	Min. :1350	Min. :148.0	Min. : 55.0	1 :152	asian : 24	Min. :15.00	0: 3	Min. :54.0	Min. : 87.0	asian : 25	Min. :18.00	2 :189	Min. :60.00	Min. :110.0	0: 2	2 :101	never :282	0 :282	0 :282
1st Qu.:5426	1st Qu.:1469	1st Qu.:272.2	1st Qu.:108.2	0 :145	black :131	1st Qu.:23.00	1: 85	1st Qu.:62.0	1st Qu.:115.0	black :132	1st Qu.:25.00	5 :165	1st Qu.:68.00	1st Qu.:155.0	1:600	1 :100	now :216	1 :216	5 :101
Median :6907	Median :1574	Median :280.0	Median :120.0	2 :117	mexican: 18	Median :27.00	2:229	Median :64.0	Median :125.0	mexican: 17	Median :30.00	4 :131	Median :71.00	Median :170.0	2: 6	3 :100	once_not_now : 60	2 : 52	1 : 75
Mean :6090	Mean :1559	Mean :278.8	Mean :119.3	3 : 82	mixed : 14	Mean :27.54	3: 32	Mean :64.1	Mean :128.9	mixed : 15	Mean :30.44	1 : 91	Mean :70.27	Mean :170.6	3: 1	7 : 90	until_pregnancy: 52	3 : 15	2 : 73
3rd Qu.:7792	3rd Qu.:1650	3rd Qu.:288.0	3rd Qu.:131.0	4 : 49	white :423	3rd Qu.:31.00	4:152	3rd Qu.:66.0	3rd Qu.:140.0	white :421	3rd Qu.:35.00	3 : 20	3rd Qu.:72.00	3rd Qu.:185.0	5: 1	4 : 71	NA	4 : 13	3 : 40
Max. :9263	Max. :1714	Max. :338.0	Max. :174.0	5 : 31	NA	Max. :43.00	5:109	Max. :72.0	Max. :220.0	NA	Max. :53.00	0 : 12	Max. :78.00	Max. :260.0	NA	5 : 65	NA	7 : 13	6 : 17
NA	NA	NA	NA	(Other): 34	NA	NA	NA	NA	NA	NA	NA	(Other): 2	NA	NA	NA	(Other): 83	NA	(Other): 19	(Other): 22

Mom’s smoking and baby’s weight

ggplot(smoking) +
  geom_boxplot(aes(x = smoke, y = weight))

Mom’s smoking and baby’s weight with reordered x-axis

ggplot(smoking) +
  geom_boxplot(aes(x = reorder(smoke, weight, FUN = median), y = weight))

Mom’s race and baby’s weight

ggplot(smoking) +
  geom_boxplot(aes(x = reorder(mom.race, weight, FUN = median), y = weight))

Dad’s race and baby’s weight

ggplot(smoking) +
  geom_boxplot(aes(x = reorder(dad.race, weight, FUN = median), y = weight))

Mom’s race and baby’s weight and dad’s race

ggplot(smoking) +
  geom_boxplot(aes(x = reorder(mom.race, weight, FUN = median), y = weight)) +
  geom_jitter(aes(x = reorder(mom.race, weight, FUN = median), y = weight, color = dad.race), alpha = 0.5, width = 0.3)

Mom’s height and moms’s weight

ggplot(smoking) +
  geom_point(aes(x = mom.height, y = mom.weight, color = mom.race), alpha = 0.5) +
  geom_smooth(aes(x = mom.height, y = mom.weight), method = "lm")

model = lm (data = smoking, formula = mom.weight ~ mom.height)
summary(model)


Call:
lm(formula = mom.weight ~ mom.height, data = smoking)

Residuals:
    Min      1Q  Median      3Q     Max 
-38.579 -11.933  -3.515   7.276  94.839 

Coefficients:
            Estimate Std. Error t value Pr(>|t|)    
(Intercept)  -86.174     18.671  -4.615 4.79e-06 ***
mom.height     3.354      0.291  11.526  < 2e-16 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 18.55 on 608 degrees of freedom
Multiple R-squared:  0.1793,    Adjusted R-squared:  0.178 
F-statistic: 132.8 on 1 and 608 DF,  p-value: < 2.2e-16

Dad’s height and dad’s weight

ggplot(smoking) +
  geom_point(aes(x = dad.height, y = dad.weight, color = mom.race), alpha = 0.5) +
  geom_smooth(aes(x = dad.height, y = dad.weight), method = "lm")

model = lm (data = smoking, formula = dad.weight ~ dad.height)
summary(model)


Call:
lm(formula = dad.weight ~ dad.height, data = smoking)

Residuals:
    Min      1Q  Median      3Q     Max 
-48.067 -13.067  -1.825  10.554  86.243 

Coefficients:
             Estimate Std. Error t value Pr(>|t|)    
(Intercept) -132.2898    18.8057  -7.035  5.4e-12 ***
dad.height     4.3105     0.2674  16.120  < 2e-16 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 19.02 on 608 degrees of freedom
Multiple R-squared:  0.2994,    Adjusted R-squared:  0.2983 
F-statistic: 259.9 on 1 and 608 DF,  p-value: < 2.2e-16

Mom’s weight and dad’s weight

ggplot(smoking) +
  geom_point(aes(x = mom.weight, y = dad.weight, color = mom.race, shape = dad.race), alpha = 0.5)

model = lm (data = smoking, formula = dad.weight ~ mom.weight)
summary(model)


Call:
lm(formula = dad.weight ~ mom.weight, data = smoking)

Residuals:
    Min      1Q  Median      3Q     Max 
-57.481 -15.817  -2.051  14.097  93.646 

Coefficients:
             Estimate Std. Error t value Pr(>|t|)    
(Intercept) 141.54810    5.74900  24.621  < 2e-16 ***
mom.weight    0.22551    0.04407   5.117 4.16e-07 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 22.25 on 608 degrees of freedom
Multiple R-squared:  0.04129,   Adjusted R-squared:  0.03972 
F-statistic: 26.19 on 1 and 608 DF,  p-value: 4.159e-07

Mom’s smoking and mom’s education

ggplot(smoking) +
  geom_bar(aes(x = mom.edu, fill = smoke), position = "fill")

ggplot(smoking) +
  geom_bar(aes(x = factor(mom.edu), fill = smoke), position = "fill")

Mom’s smoking and the family’s income

ggplot(smoking) +
  geom_bar(aes(x = factor(income), fill = smoke), position = "fill")

Mom’s race and mom’s weight

ggplot(smoking) +
  geom_boxplot(aes(x = reorder(mom.race, mom.weight, FUN = median), y = mom.weight))

Dad’s race and dad’s weight

ggplot(smoking) +
  geom_boxplot(aes(x = reorder(dad.race, dad.weight, FUN = median), y = dad.weight))

Data Visualization

Agenda

Setting up

Titanic Survival

Load and explore the dataset

Survival by class

Survival by sex

Survival by age

Survival by age & sex

Survival by class & age

Smoking and Pregnancy

Load and explore the dataset

Mom’s smoking and baby’s weight

Mom’s smoking and baby’s weight with reordered x-axis

Mom’s race and baby’s weight

Dad’s race and baby’s weight

Mom’s race and baby’s weight and dad’s race

Mom’s height and moms’s weight

Dad’s height and dad’s weight

Mom’s weight and dad’s weight

Mom’s smoking and mom’s education

Mom’s smoking and the family’s income

Mom’s race and mom’s weight

Dad’s race and dad’s weight

The End