Linear-Regression-in-R/code at main · Steveximo/Linear-Regression-in-R · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
# Importing data and loading packages
birthweight <- rio::import(here::here("Birth_baby_weight.csv"))
 if(!require(pacman)) install.packages("pacman")
pacman::p_load( gt,car, dplyr, lessR, ggplot2, magrittr, janitor, plotrix, broom,kableExtra )

# Data manipulation
 bweight <- birthweight %>%
    mutate(birthwt = bwt_ounce*0.02835,
                       mheight = mheight_inches*0.0254,
                       mweight = mweight_pounds*0.4536)
 bweight <- bweight %>%
    select(Gender,birthwt, gestation_days,mother_age, mweight) %>%
    filter(gestation_days > 200 & gestation_days < 350)

head(bweight,5)

any(is.na(bweight))

 dim(bweight)

 bweight$Gender <- as.factor(bweight$Gender)

levels(bweight$Gender) <- c("male","female")

 str(bweight)


# Descriptive statistics of the data
 summary(bweight)

 table1 <- table(bweight$Gender)

 piepercent <-paste0(round(100 * table1/sum(table1)), "%")

plotrix::pie3D(table1,radius = 1.5,
                                 explode = 0.25,
                                 labels = piepercent,
                                  col = c("purple","orange"),
                                  main = "Births according to sex  ",
                                  col.main="blue")
 legend("topright",
                c("males","females"),
                cex = 0.8,
                fill =c("purple","orange") )


 table2 <- table(bweight$smoking_status)

 barplot(table2,
                    main ="Mother's smoking status",
                    ylab = "Number of mothers",
                    col.main = "blue",
                    col.lab = "darkblue",
                    col = c("lightgreen","red") )

# Correlation and Linear regression analysis
 ggplot(data = bweight,
                  mapping=aes(x=gestation_days,
                                                   y=birthwt))+
   geom_point(col='green',
                 +              size=2)+
   geom_smooth(method='lm',
                  +               col= 'blue')+
   labs(title="Baby's birth weight Vs Gestation period",
                  x="Gestation period",
                  y="Birth Weight")

  cor.test(bweight$birthwt,
                      bweight$gestation_days,method = "pearson")

 # fitting a simple linear model
model1 <- lm(bweight$birthwt~bweight$gestation_days)
 summary(model1)
hist(residuals(model1))
 ggplot(data = bweight,
      mapping=aes(x=mweight,
                  y=birthwt))+
  geom_point(col='green',
             size=2)+
  labs(title="Baby's birth weight Vs mother weight",
       x="Mother's weight",
       y="Birth Weight")

cor.test(bweight$birthwt,
         bweight$mweight,method = "pearson")


 model2 <- lm(bweight$birthwt~bweight$mweight)

 summary(model2)

plot(bweight$birthwt,residuals(model2))
 hist(residuals(model2))

cor.test(bweight$birthwt,
                      bweight$mweight,method = "pearson")

shapiro.test(bweight$birthwt)

car::leveneTest(bweight$birthwt ~ bweight$Gender)


 t.test(data = bweight, birthwt ~ Gender,
                 alt = "two.sided", var.eq = T)


 boxplot(birthwt ~ Gender, data = bweight,
                    main="BOX PLOT OF BIRTH WEIGHT BY GENDER",
                         ylab = "Birth weight(kg)")

# Fitting a multiple linear regression model
 model3 <- lm(birthwt ~ mweight + gestation_days,
             data = bweight)
summary(model3)

 model <- lm(birthwt ~ mweight + gestation_days + Gender,
             data = bweight)
 summary(model)