MSP:

devtools::load_all()
data("BRFSS19952010")
library(dplyr)
library(ggplot2)
library(lubridate)

model_1 <- specification("MEDCOST", model = "DID")
model_3 <- specification("MEDCOST", model = "event_time")

BRFSS_lowIncome <- filter(BRFSS19952010, INCOME %in% c("1", "2", "3", "4"))
BRFSS_highIncome <- filter(BRFSS19952010, INCOME %in% c("4", "5", "6", "7", "8"))


fit2 <- regression_analysis(BRFSS_lowIncome, model_1, cluster = BRFSS_lowIncome$state)
fit5 <- regression_analysis(BRFSS_lowIncome, model_3, cluster = BRFSS_lowIncome$state, event_plot = TRUE)

fit2 <- regression_analysis(BRFSS_highIncome, model_1, cluster = BRFSS_highIncome$state)
fit5 <- regression_analysis(BRFSS_highIncome, model_3, cluster = BRFSS_highIncome$state, event_plot = TRUE)


fit2 <- subgroup_regression("SEX=='1'", BRFSS19952010, model_1, cluster = "state")
fit5 <- subgroup_regression("SEX=='1'", BRFSS19952010, model_3, cluster = "state", event_plot = TRUE)

fit2 <- subgroup_regression("SEX=='2'", BRFSS19952010, model_1, cluster = "state")
fit5 <- subgroup_regression("SEX=='2'", BRFSS19952010, model_3, cluster = "state", event_plot = TRUE)


fit2 <- subgroup_regression("EDUCA %in% c('1', '2', '3', '4')", BRFSS19952010, model_1, cluster = "state")
fit5 <- subgroup_regression("EDUCA %in% c('1', '2', '3', '4')", BRFSS19952010, model_3, cluster = "state", event_plot = TRUE)

fit2 <- subgroup_regression("EDUCA %in% c('5', '6')", BRFSS19952010, model_1, cluster = "state")
fit5 <- subgroup_regression("EDUCA %in% c('5', '6')", BRFSS19952010, model_3, cluster = "state", event_plot = TRUE)



BRFSS %>%
  mutate(year = as.numeric(year), treated = factor(treated)) %>%
  group_by(year, treated) %>%
  summarize(MEDCOST = mean(MEDCOST, na.rm = TRUE)) %>%
  ggplot(aes(year, MEDCOST, color = treated)) +
  geom_line()

paste(BRFSS$year, BRFSS$IMONTH, 1, sep="-") %>%
  ymd() %>%
  format("%Y-%m")


BRFSS %>%
  mutate(treated = factor(treated),
         month = as.numeric(year)*12+IMONTH
         ) %>%
  group_by(month, treated) %>%
  summarize(MEDCOST = mean(MEDCOST, na.rm = TRUE)) %>%
  ggplot(aes(month, MEDCOST, color = treated)) +
  geom_line()


model_cps <- specification("HIMCAID", "DID")
model_cps <- specification("HIMCAID", "event_time")
fit_cps <- regression_analysis(CPS, model_cps, cluster="state")
fit_cps <- regression_analysis(CPS, model_cps, cluster="state", event_plot = TRUE)

CPS %>%
  mutate(year = as.numeric(year), treated = factor(treated)) %>%
  group_by(year, treated) %>%
  summarize(HIMCARE = mean(HIMCARE, na.rm = TRUE)) %>%
  ggplot(aes(year, HIMCARE, color = treated)) +
  geom_line()