Lab 07

Author

Affiliation

Moisieiev Vasyl

Kyiv School of Economics

Import

library(tidyverse)
library(tidymodels)
library(DAAG)

allbacks

Exercise 1

ggplot(allbacks, aes(x = volume, y = weight)) +
  geom_point() +                                
  geom_smooth(method = "lm", se = FALSE, color = "blue") + 
  labs(title = "Relationship between Volume and Weight",
       x = "Volume",
       y = "Weight") +
  theme_minimal()

Exercise 2 + 3

weight_fit <- lm(weight ~ volume, data = allbacks)

tidy(weight_fit)

Intercept is 107.68 and the slope is 0.7086. It means that for each additional unit of volume, the weight of the allbacks increases by 0.7086 units.

Exercise 4

summary(weight_fit)$r.squared

[1] 0.8026346

The R-squared value is 0.80263, it means that 80.26% of the variation in weight can be explained by the volume of the allbacks.

Exercise 5

ggplot(allbacks, aes(x = volume, y = weight, color = cover, shape = cover)) +
  geom_point(size = 3) +  
  geom_smooth(method = "lm", se = FALSE, aes(group = cover)) +  
  geom_smooth(method = "lm", se = FALSE, color = "gray", linetype = "dashed") + 
  labs(
    title = "Relationship Between Volume and Weight by Cover Type",
    x = "Volume",
    y = "Weight",
    color = "Cover Type",
    shape = "Cover Type"
  ) +
  theme_minimal()

Exercise 6 + 7 + 8

weight_cover_fit <- lm(weight ~ volume + cover, data = allbacks)

tidy(weight_cover_fit)

Because coverhb service as a baseline category. And coverpb showes how it differs from the coverhb.

Intercept in this case is a baseline and slope showes how it differs from the baseline by each unit of volume.

Exercise 9

summary(weight_cover_fit)$r.squared

[1] 0.9274776

It means that 0.9274 this prediction will be right in 92.74% of the cases.

Exercise 10

Second model is better because it has higher R-squared value. It means that 92.74% of the variation in weight can be explained by the volume and cover type of the allbacks.

Exercise 11

new_data <- data.frame(volume = 1000, cover = "hb")  
predicted_weight <- predict(weight_cover_fit, newdata = new_data)

predicted_weight

       1 
915.9166