Statistical Modelling and Variable Selection in Climate Science
363
Table 1 Data on rainfall, wind speed, precipitation and relative humidity
Observation
number (i)
Rainfall (y i ) Wind speed (x 1i ) Precipitation(x 2i ) Relative humidity
(x 3i )
1
122.0
18.3
29.1
52.0
2
114.6
15.7
23.5
50.8
3
135.4
15.7
29.8
65.2
4
136.1
19.5
25.5
64.0
5
135.7
15.4
25.7
68.7
6
119.1
19.7
20.2
50.2
7
128.1
19.2
26.4
57.5
8
126.1
15.1
20.6
64.3
9
125.2
19.6
24.2
53.5
10
137.4
18.2
26.5
67.2
We use the data in R software (available at https://www.r-project.org/) to find the
relationship of rainfall with wind speed, precipitation, and relative humidity. The
variables on rainfall, wind speed, precipitation, and relative humidity are denoted as
rain, windspeed, precipitation, and relhum respectively. The data on
the variables in R software is entered as follows:
windspeed=c(18.3, 15.7, 15.7, 19.5, 15.4, 19.7, 19.2,
15.1, 19.6, 18.2)
precipitation=c(29.1, 23.5, 29.8, 25.5, 25.7, 20.2,
26.4, 20.6, 24.2, 26.5)
relhum=c(52.0, 50.8, 65.2, 64.0, 68.7, 50.2, 57.5,
64.3, 53.5, 67.2)
rain=c(122.0, 114.6, 135.4, 136.1, 135.7, 119.1, 128.1,
126.1, 125.2, 137.4)
The R command to fit a linear model is lm(). The R command
summary(lm(rain ~ windspeed + precipitation + relhum))
is used to fit the multiple linear regression model of rain on windspeed,
precipitation, and relhum. The outcome of R software is copied from the
console of software and reproduced as follows:
363
Table 1 Data on rainfall, wind speed, precipitation and relative humidity
Observation
number (i)
Rainfall (y i ) Wind speed (x 1i ) Precipitation(x 2i ) Relative humidity
(x 3i )
1
122.0
18.3
29.1
52.0
2
114.6
15.7
23.5
50.8
3
135.4
15.7
29.8
65.2
4
136.1
19.5
25.5
64.0
5
135.7
15.4
25.7
68.7
6
119.1
19.7
20.2
50.2
7
128.1
19.2
26.4
57.5
8
126.1
15.1
20.6
64.3
9
125.2
19.6
24.2
53.5
10
137.4
18.2
26.5
67.2
We use the data in R software (available at https://www.r-project.org/) to find the
relationship of rainfall with wind speed, precipitation, and relative humidity. The
variables on rainfall, wind speed, precipitation, and relative humidity are denoted as
rain, windspeed, precipitation, and relhum respectively. The data on
the variables in R software is entered as follows:
windspeed=c(18.3, 15.7, 15.7, 19.5, 15.4, 19.7, 19.2,
15.1, 19.6, 18.2)
precipitation=c(29.1, 23.5, 29.8, 25.5, 25.7, 20.2,
26.4, 20.6, 24.2, 26.5)
relhum=c(52.0, 50.8, 65.2, 64.0, 68.7, 50.2, 57.5,
64.3, 53.5, 67.2)
rain=c(122.0, 114.6, 135.4, 136.1, 135.7, 119.1, 128.1,
126.1, 125.2, 137.4)
The R command to fit a linear model is lm(). The R command
summary(lm(rain ~ windspeed + precipitation + relhum))
is used to fit the multiple linear regression model of rain on windspeed,
precipitation, and relhum. The outcome of R software is copied from the
console of software and reproduced as follows: