fred_inf.R

library(fredr)
library(tidyverse)
library(ggdirectlabel) # remotes::install_github("mattcowgill/ggdirectlabel")

# You must have a FRED API key; 
# see http://sboysel.github.io/fredr/articles/fredr.html

us_cpi <- fredr("CPIAUCSL")

us_sent <- fredr("UMCSENT") |> 
  select(date, sentiment = value) |> 
  filter(!is.na(sentiment))

my_theme <- theme_minimal(base_family = "Roboto Condensed", 
                          base_size = 16) %+replace%
  theme(plot.title.position = "plot",
        plot.caption.position = "plot",
        plot.caption = element_text(hjust = 0,
                                    size = rel(0.5)),
        plot.title = ggtext::element_textbox_simple(
          size = rel(1.25),
          hjust = 0,
          vjust = 1,
          face = "bold"
        ),
        plot.subtitle = ggtext::element_textbox_simple(
          colour = "grey30",
          size = rel(1),
          hjust = 0,
          vjust = 1
        ))

delta_inf <- us_cpi |> 
  select(date, cpi = value) |> 
  arrange(date) |> 
  mutate(inf = (cpi / lag(cpi, 12L, order_by = date)) - 1,
         ave_inf = slider::slide_mean(inf, before = 119L, complete = TRUE),
         delta_inf = inf - ave_inf) |> 
  filter(!is.na(delta_inf)) 

comb <- delta_inf |> 
  right_join(us_sent,
             by = 'date') |> 
  mutate(decade = paste0(str_sub(format(date, "%Y"), 1, 3),
                         "0s")) 

plot_inf <- comb |> 
  select(date, 
         `Average inf. over prior 10 years` = ave_inf, 
         `Current inflation` = inf) |> 
  pivot_longer(!date) |> 
  ggplot(aes(x = date, y = value, col = name)) +
  geom_line() +
  geom_richlegend(aes(label = name),
                  family = "Roboto Condensed",
                  size = 14 / .pt) +
  my_theme +
  scale_y_continuous(labels = scales::percent) +
  scale_colour_brewer(palette="Set1") +
  theme(legend.position = "none",
        axis.title = element_blank()) +
  labs(subtitle = "Current inflation and a rolling 10 year average (United States)",
       caption = "Source: FRED")

plot_inf_diff <- comb |> 
  ggplot(aes(x = date, y = delta_inf)) +
  geom_hline(yintercept = 0) +
  geom_line(colour = "purple") +
  my_theme +
  scale_y_continuous(labels = scales::percent) +
  theme(axis.title = element_blank()) +
  labs(subtitle = "Difference between current inflation and rolling 10 year average (percentage points)",
       caption = "Source: FRED")

plot_scatter <- comb |> 
  rename(Decade = decade) |> 
  ggplot(aes(x = delta_inf,
             y = sentiment,
             col = Decade)) +
  geom_point(alpha = 0.5) +
  ggrepel::geom_text_repel(data = ~mutate(., 
                                          label = if_else(date == max(date),
                                                          format(date, "%B\n%Y"),
                                                          "")),
                           aes(label = label),
                           min.segment.length = 0,
                           lineheight = 0.9,
                           nudge_x = -.01,
                           nudge_y = -10,
                           max.overlaps = 10000,
                           show.legend = FALSE) +
  my_theme +
  labs(x = "Difference between current inflation\nand average over prior 10 years (ppts)",
       y = "Consumer sentiment",
       caption = "Source: FRED")

my_save <- function(filename, plot) {
  ggsave(filename, plot,
         width = 20, height = 13, units = "cm", dpi = "retina", bg = "white")
}

my_save("plot_inf.png", plot_inf)
my_save("plot_inf_diff.png", plot_inf_diff)
my_save("plot_scatter.png", plot_scatter)