library(ggplot2)
library(tidyverse)

fedData <- read.csv("PATH TO CSV FILE", header = TRUE)

fedDataReduced <-
  fedData %>%
  transmute(date = `ï..Date`,
            recessionProb = Rec_prob,
            inRecession = NBER_Rec) %>%
  filter(!is.na(recessionProb) & !is.na(inRecession)) %>%
  mutate(date = as.POSIXct(strptime(date,format="%m/%d/%Y"))) %>%
  arrange(date) %>%
  mutate(newRecession = ifelse(lag(inRecession, default = 0) == 0 & inRecession == 1, 1, 0)) %>%
  mutate(leadR1 = lead(inRecession,1, default = 0),
         leadR2 = lead(inRecession,2, default = 0),
         leadR3 = lead(inRecession,3, default = 0),
         leadR4 = lead(inRecession,4, default = 0),
         leadR5 = lead(inRecession,5, default = 0),
         leadR6 = lead(inRecession,6, default = 0),
         leadR7 = lead(inRecession,7, default = 0),
         leadR8 = lead(inRecession,8, default = 0),
         leadR9 = lead(inRecession,9, default = 0),
         leadR10 = lead(inRecession,10, default = 0),
         leadR11 = lead(inRecession,11, default = 0),
         leadR12 = lead(inRecession,12, default = 0),
         sumR = leadR1+leadR2+leadR3+leadR4+leadR5+leadR6+leadR7+leadR8+leadR9+leadR10+leadR11+leadR12,
         recessionNextYear = ifelse(sumR > 0 , 1, 0)) %>%
  mutate(leadNR1 = lead(newRecession,1, default = 0),
         leadNR2 = lead(newRecession,2, default = 0),
         leadNR3 = lead(newRecession,3, default = 0),
         leadNR4 = lead(newRecession,4, default = 0),
         leadNR5 = lead(newRecession,5, default = 0),
         leadNR6 = lead(newRecession,6, default = 0),
         leadNR7 = lead(newRecession,7, default = 0),
         leadNR8 = lead(newRecession,8, default = 0),
         leadNR9 = lead(newRecession,9, default = 0),
         leadNR10 = lead(newRecession,10, default = 0),
         leadNR11 = lead(newRecession,11, default = 0),
         leadNR12 = lead(newRecession,12, default = 0),
         sumNR = leadNR1+leadNR2+leadNR3+leadNR4+leadNR5+leadNR6+leadNR7+leadNR8+leadNR9+leadNR10+leadNR11+leadNR12,
         newRecessionNextYear = ifelse(sumNR > 0 , 1, 0)) %>%
  mutate(bucket = ifelse(recessionProb < 1, 9, NA)) %>%
  mutate(bucket = ifelse(recessionProb < 0.9, 8, bucket)) %>%
  mutate(bucket = ifelse(recessionProb < 0.8, 7, bucket)) %>%
  mutate(bucket = ifelse(recessionProb < 0.7, 6, bucket)) %>%
  mutate(bucket = ifelse(recessionProb < 0.6, 5, bucket)) %>%
  mutate(bucket = ifelse(recessionProb < 0.5, 4, bucket)) %>%
  mutate(bucket = ifelse(recessionProb < 0.4, 3, bucket)) %>%
  mutate(bucket = ifelse(recessionProb < 0.3, 2, bucket)) %>%
  mutate(bucket = ifelse(recessionProb < 0.2, 1, bucket)) %>%
  mutate(bucket = ifelse(recessionProb < 0.1, 0, bucket)) %>%
  group_by(bucket) %>%
  summarize(n = n(),
            recCount = sum(recessionNextYear),
            nRecCount = sum(newRecessionNextYear),
            percentRec = recCount / n,
            percentNRec = nRecCount / n) %>%
  mutate(pred = bucket/10,
         dataCount = ifelse(n < 150, 150, 400)) %>%
  mutate(dataCount = ifelse(n < 100, 100, dataCount)) %>%
  mutate(dataCount = ifelse(n < 50, 50, dataCount)) %>%
  mutate(dataCount = ifelse(n < 25, 25, dataCount)) %>%
  mutate(dataCount = ifelse(n < 10, 10, dataCount))

ggplot(fedDataReduced) +
  geom_line(aes(x=pred,y=pred)) +
  geom_line(aes(x=pred,y=pred - 0.1)) +
  geom_point(aes(x=pred,y=percentRec, colour = as.factor(dataCount))) +
  labs(x="Prediction",
       y="Occurence Rate",
       colour="Data Count Less Than")

ggplot(fedDataReduced) +
  geom_line(aes(x=pred,y=pred)) +
  geom_line(aes(x=pred,y=pred - 0.1)) +
  geom_point(aes(x=pred,y=percentNRec, colour = as.factor(dataCount))) +
  labs(x="Prediction",
       y="Occurence Rate",
       colour="Data Count Less Than")