Plotting parallel regression lines

model-parallel-lines.R

# Fit model with parallel lines (main effects only)
m_pr <- lm(log(price) ~ Surface + factor(artistliving), data = pp_Surf_lt_5000)

## Option 1

# Save regression coefficients in a tibble
m_pr_coefs <- m_pr %>%
  tidy() %>%
  select(term, estimate)

# Save regression coefficients separately as numeric values
deceased_int <- m_pr_coefs %>% 
  filter(term == "(Intercept)") %>%
  select(estimate) %>% 
  pull()
living_int <- m_pr_coefs %>% 
  filter(term == "factor(artistliving)1" | term == "(Intercept)") %>%
  select(estimate) %>% 
  pull() %>%
  sum()
surf_slope <- m_pr_coefs %>% 
  filter(term == "Surface") %>%
  select(estimate) %>% 
  pull()

# Plot lines with geom_abline
ggplot(data = pp_Surf_lt_5000,
       aes(y = log(price), x = Surface, color = factor(artistliving))) +
  geom_point(alpha = 0.3) +
  geom_abline(intercept = deceased_int, slope = surf_slope, color = "#F57670", lwd = 1) +
  geom_abline(intercept = living_int, slope = surf_slope, color = "#1FBEC3", lwd = 1) +
  labs(x = "Surface", y = "Log(price)", color = "Living artist")

## Option 2

# Augment the data with model results, including fitted values
m_pr_aug <- augment(m_pr)

# Plot lines with geom_line for predicted values
ggplot(data = m_pr_aug, mapping = aes(y = log.price., x = Surface, color = factor.artistliving.)) +
  geom_point(alpha = 0.3) +
  geom_line(aes(y = .fitted), lwd = 1)