Problem 1: Regression Analysis
The following dataset on body weight and on futher observations on a number of animals is given.
1 |
176 |
471 |
5.0 |
161 |
2 |
177 |
463 |
4.2 |
121 |
3 |
178 |
481 |
4.9 |
157 |
4 |
179 |
470 |
3.0 |
165 |
5 |
179 |
496 |
6.8 |
136 |
6 |
180 |
491 |
4.9 |
123 |
7 |
181 |
518 |
4.4 |
163 |
8 |
182 |
511 |
4.4 |
149 |
9 |
183 |
510 |
3.5 |
143 |
10 |
184 |
541 |
4.7 |
130 |
The data can be read from https://charlotte-ngs.github.io/asmss2022/data/asm_bw_mult_reg.csv. The additional columns contain data on body condition score (BCS) and height (HEI).
Tasks
- Build a regression model of body weight on the other observations using the dataset given above.
- Set up the matrix \(\mathbf{X}\) and the vectors \(\mathbf{y}\), \(\mathbf{b}\) and \(\mathbf{e}\).
- Compute estimate for the regression coefficients in the model defined above.
Your Solution
- Start by building the model which includes the decision of what type of components go into the vector \(\mathbf{b}\). Then write down the model in matrix-vector notation.
- Determine all the components of \(\mathbf{X}\), \(\mathbf{y}\), \(\mathbf{b}\) and \(\mathbf{e}\).
- Compute the solution for $
Problem 2
Use the same dataset as in Problem 1 and verify your results using the function lm()
in R.
Your Solution
- Read the data into a dataframe
- Use the
lm()
function which takes a formula to define the model and a dataframe
- Use the function
summary()
on the result of lm()
to show the results
Latest Changes: 2022-02-26 16:36:25 (pvr)
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