library(tidyverse)
entest_fp <-"/mmfs1/gscratch/comdata/users/mjilg/mw-repo-lifecycles/case1/en-testing_0217_extensions_ve_weekly_commit_count_data.csv"
entest_df <- read.csv(entest_fp, header = TRUE) |> mutate(rd_event = "en-testing")

widetest_fp <-"/mmfs1/gscratch/comdata/users/mjilg/mw-repo-lifecycles/case1/wide-testing_0217_extensions_ve_weekly_commit_count_data.csv"
widetest_df <- read.csv(widetest_fp, header = TRUE) |> mutate(rd_event = "wide-testing")

event_fp <-"/mmfs1/gscratch/comdata/users/mjilg/mw-repo-lifecycles/case1/event_0217_extensions_ve_weekly_commit_count_data.csv"
event_df <- read.csv(event_fp, header = TRUE) |> mutate(rd_event = "default")

input_df <- bind_rows(entest_df, widetest_df, event_df)
#input_df <- bind_rows(entest_df, widetest_df)

input_df <- input_df |>
  mutate(nonbot_commit_count = commit_count - bot_commit_count)|>
  mutate(other_commit_count = nonbot_commit_count - mediawiki_dev_commit_count - wikia_commit_count - wikimedia_commit_count) |>
  mutate(wikimedia_commit_count = wikimedia_commit_count + mediawiki_dev_commit_count + wikia_commit_count) |>
  dplyr::select(-mediawiki_dev_commit_count) |>
  dplyr::select(-wikia_commit_count)

library(MASS)

library(lme4)
library(dplyr)

#get into mlm format
long_df <- input_df |>
  pivot_longer(cols = c(other_commit_count, wikimedia_commit_count),
               names_to = "commit_type",
               values_to = "lengthened_commit_count")

intermediate_long_df <- long_df |>
  mutate(commit_share = lengthened_commit_count / (nonbot_commit_count)) |>
  mutate(log_commits = log1p(lengthened_commit_count))|>
  mutate(scaled_long_commits = lengthened_commit_count / 10) 

library(rdd)

intermediate_long_df <- intermediate_long_df |>
  drop_na()

var(intermediate_long_df$commit_share) # 1253.343
mean(intermediate_long_df$commit_share) # 44.92381
median(intermediate_long_df$commit_share) # 39.5

get_optimal_bandwidth <- function(df){
  bw <- tryCatch({
    IKbandwidth(df$relative_week, df$commit_share, cutpoint = 0, verbose = FALSE, kernel = "triangular")
  }, error = function(e) {
    NA
  })
}

optimal_bandwidth <- get_optimal_bandwidth(intermediate_long_df)


window_num <- 18
final_long_df <- intermediate_long_df |>
  filter(relative_week >= (- window_num) & relative_week <= (window_num)) 

library(fitdistrplus)
descdist(long_df$lengthened_commit_count, discrete=FALSE)
#start_values <- list(shape1 = 1, shape2 = 1)
#fit <- MASS::fitdistr(as.numeric(long_df$lengthened_commit_count), "negative binomial")
print(fit)

#NOTE should not run if you've already dropped NA
mlm <- glmer.nb(lengthened_commit_count ~ before_after*relative_week +
                  (before_after*relative_week|commit_type) + 
                  (before_after*relative_week|rd_event),
                control=glmerControl(optimizer="bobyqa",
                                     optCtrl=list(maxfun=2e5)), nAGQ=0,
                data=long_df)
#mlm <- lmer(lengthened_commit_count ~ before_after*relative_week+
#                    (before_after*relative_week|commit_type) + 
#                    (before_after*relative_week|rd_event) ,data=long_df)
summary(mlm)
qqnorm(residuals(mlm))
res <- ranef(mlm)
print(res)

#final_long_df <- final_long_df |>
#  drop_na()
library(performance)
#descdist(long_df$commit_share, discrete=FALSE)
#fit <- MASS::fitdistr(as.numeric(long_df$commit_share), "normal")
#print(fit)
wikimedia_long_df <- final_long_df |>
  filter(commit_type == "wikimedia_commit_count")
wikimedia_share_lmer <- lmer(commit_share ~ before_after*relative_week +
                               (1| rd_event),
                  data=wikimedia_long_df)
summary(wikimedia_share_lmer)
icc(wikimedia_share_lmer)

other_long_df <- final_long_df |>
  filter(commit_type == "other_commit_count")
other_share_lmer <- lm(commit_share ~ before_after*relative_week,
                           data=other_long_df)
summary(other_share_lmer)
icc(other_share_lmer)

#power analysis 
#library(simr)
#simrOptions(progress=FALSE)

## Intercept and slopes for intervention, time1, time2, intervention:time1, intervention:time2
#wmf_fixed <- c(0.511, -0.166, 0.002, 0.007)
## Random intercepts for participants clustered by class
#wmf_rand <- matrix(c(
#  0.01,  0.005, 0.002, 0.001,
#  0.005, 0.02,  0.003, 0.004,
#  0.002, 0.003, 0.015, 0.006,
#  0.001, 0.004, 0.006, 0.01
#), nrow=4, byrow=TRUE)
## residual variance
#wmf_res <- 0.2065

#wmf_model <- makeLmer(commit_share ~ before_after*relative_week + (before_after*relative_week | rd_event),
#                      fixef=wmf_fixed, VarCorr=wmf_rand, sigma=wmf_res, data=wikimedia_long_df)

#sim_treat <- powerSim(wmf_model, nsim=100, test = fcompare(commit_share~relative_week))
#sim_treat

#model_ext_subj <- extend(wmf_model, within="rd_event+before_after+relative_week", n=30)
#sim_treat_subj <- powerSim(model_ext_subj, nsim=100, test = fcompare(commit_share~before_after*relative_week))
#sim_treat_subj

#p_curve_treat <- powerCurve(model_ext_subj, test=fcompare(commit_share~before_after*relative_week), 
#                            within="rd_event+before_after+relative_week", 
#                            breaks=c(5,10,15,20))
#plot(p_curve_treat)