From c8313a904fcbc04f0a375b092158edfb2afcab24 Mon Sep 17 00:00:00 2001 From: mjgaughan Date: Tue, 7 May 2024 18:40:38 -0500 Subject: [PATCH] updating rdd scripts w pop level --- R/.Rhistory | 806 +++++++++--------- R/contribRDDAnalysis.R | 11 +- R/popRDDAnalyssis.R | 55 ++ R/readmeRDDAnalysis.R | 17 +- .../deb_contrib_pop_change.csv | 0 .../deb_readme_pop_change.csv | 0 6 files changed, 473 insertions(+), 416 deletions(-) create mode 100644 R/popRDDAnalyssis.R rename 042724_draft_deb_contrib_person.csv => final_data/deb_contrib_pop_change.csv (100%) rename 42524_new_readme.csv => final_data/deb_readme_pop_change.csv (100%) diff --git a/R/.Rhistory b/R/.Rhistory index 74b9299..bc75eda 100644 --- a/R/.Rhistory +++ b/R/.Rhistory @@ -1,233 +1,11 @@ -filter(week >= (27 - window_num) & week <= (27 + window_num)) |> -mutate(D = ifelse(week > 27, 1, 0)) -windowed_sample_data$scaled_project_age <- scale(windowed_sample_data$age_of_project) -windowed_sample_data$week_offset <- windowed_sample_data$week - 27 -all_actions_sample_data <- windowed_sample_data[which(windowed_sample_data$observation_type == "all"),] -#test model -test_model <- lmer(count ~ D * I(week_offset) + scaled_project_age + (D * I(week_offset)|upstream_vcs_link), data=all_actions_sample_data, REML=FALSE) -summary(test_model) -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=predict(test_model), group=upstream_vcs_link), show.legend = FALSE) + -theme_bw() -p -# https://www.bristol.ac.uk/cmm/learning/videos/random-intercepts.html#exvar -#making some random data -sampled_data <- readme_df[sample(nrow(readme_df), 220), ] -expanded_sample_data <- expand_timeseries(sampled_data[1,]) -for (i in 2:nrow(sampled_data)){ -expanded_sample_data <- rbind(expanded_sample_data, expand_timeseries(sampled_data[i,])) -} -windowed_sample_data <- expanded_sample_data |> -filter(week >= (27 - window_num) & week <= (27 + window_num)) |> -mutate(D = ifelse(week > 27, 1, 0)) -windowed_sample_data$scaled_project_age <- scale(windowed_sample_data$age_of_project) -windowed_sample_data$week_offset <- windowed_sample_data$week - 27 -all_actions_sample_data <- windowed_sample_data[which(windowed_sample_data$observation_type == "all"),] -#test model -test_model <- lmer(count ~ D * I(week_offset) + scaled_project_age + (D * I(week_offset)|upstream_vcs_link), data=all_actions_sample_data, REML=FALSE) -summary(test_model) -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=predict(test_model), group=upstream_vcs_link), show.legend = FALSE) + -theme_bw() -p -#test model -test_model <- lmer(count ~ D * I(week_offset) + scaled_project_age + (week_offset|upstream_vcs_link), data=all_actions_sample_data, REML=FALSE) -summary(test_model) -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=predict(test_model), group=upstream_vcs_link), show.legend = FALSE) + -theme_bw() -p -#test model -test_model <- lmer(count ~ D * I(week_offset) + scaled_project_age + (D * I(week_offset)|upstream_vcs_link), data=all_actions_sample_data, REML=FALSE) -summary(test_model) -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=predict(test_model), group=upstream_vcs_link), show.legend = FALSE) + -theme_bw() -p -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=predict(test_model)), show.legend = FALSE) + -theme_bw() -p -# for visualization, may have to run model for each project and then identify top 5 projects for RDD graphs -mrg_model <- lmer(count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=mrg_actions_data, REML=FALSE) -summary(mrg_model) -summary(all_model) -#test model -test_model <- lmer(count ~ D * I(week_offset) + scaled_project_age + (D * I(week_offset)||upstream_vcs_link), data=all_actions_sample_data, REML=FALSE) -summary(test_model) -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=predict(test_model)), show.legend = FALSE) + -theme_bw() -p -#test model -test_model <- lmer(count ~ D * I(week_offset) + scaled_project_age + (D * I(week_offset)|upstream_vcs_link), data=all_actions_sample_data, REML=FALSE) -summary(test_model) -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=predict(test_model)), show.legend = FALSE) + -theme_bw() -p -# https://www.bristol.ac.uk/cmm/learning/videos/random-intercepts.html#exvar -#making some random data -sampled_data <- readme_df[sample(nrow(readme_df), 220), ] -expanded_sample_data <- expand_timeseries(sampled_data[1,]) -for (i in 2:nrow(sampled_data)){ -expanded_sample_data <- rbind(expanded_sample_data, expand_timeseries(sampled_data[i,])) -} -windowed_sample_data <- expanded_sample_data |> -filter(week >= (27 - window_num) & week <= (27 + window_num)) |> -mutate(D = ifelse(week > 27, 1, 0)) -windowed_sample_data$scaled_project_age <- scale(windowed_sample_data$age_of_project) -windowed_sample_data$week_offset <- windowed_sample_data$week - 27 -all_actions_sample_data <- windowed_sample_data[which(windowed_sample_data$observation_type == "all"),] -#test model -test_model <- lmer(count ~ D * I(week_offset) + scaled_project_age + (D * I(week_offset)|upstream_vcs_link), data=all_actions_sample_data, REML=FALSE) -summary(test_model) -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=predict(test_model)), show.legend = FALSE) + -theme_bw() -p -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=mean(predict(test_model)), show.legend = FALSE)) + -theme_bw() -p -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=predict(test_model), show.legend = FALSE)) + -theme_bw() -p -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=mean(predict(test_model))), show.legend = FALSE) + -theme_bw() -p -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=median(predict(test_model))), show.legend = FALSE) + -theme_bw() -p -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=predict(test_model)), show.legend = FALSE) + -theme_bw() -p -#test model -test_model <- lmer(count ~ D * I(week_offset) + scaled_project_age + (D * I(week_offset) + scaled_project_age|upstream_vcs_link), data=all_actions_sample_data, REML=FALSE) -summary(test_model) -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=predict(test_model)), show.legend = FALSE) + -theme_bw() -p -#test model -test_model <- lmer(count ~ D * I(week_offset) + scaled_project_age + (D * I(week_offset)|upstream_vcs_link), data=all_actions_sample_data, REML=FALSE) -summary(test_model) -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=predict(test_model)), show.legend = FALSE) + -theme_bw() -p -# https://www.bristol.ac.uk/cmm/learning/videos/random-intercepts.html#exvar -#making some random data -sampled_data <- readme_df[sample(nrow(readme_df), 22), ] -expanded_sample_data <- expand_timeseries(sampled_data[1,]) -for (i in 2:nrow(sampled_data)){ -expanded_sample_data <- rbind(expanded_sample_data, expand_timeseries(sampled_data[i,])) -} -windowed_sample_data <- expanded_sample_data |> -filter(week >= (27 - window_num) & week <= (27 + window_num)) |> -mutate(D = ifelse(week > 27, 1, 0)) -windowed_sample_data$scaled_project_age <- scale(windowed_sample_data$age_of_project) -windowed_sample_data$week_offset <- windowed_sample_data$week - 27 -all_actions_sample_data <- windowed_sample_data[which(windowed_sample_data$observation_type == "all"),] -#test model -test_model <- lmer(count ~ D * I(week_offset) + scaled_project_age + (D * I(week_offset)|upstream_vcs_link), data=all_actions_sample_data, REML=FALSE) -summary(test_model) -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=predict(test_model)), show.legend = FALSE) + -theme_bw() -p -# https://www.bristol.ac.uk/cmm/learning/videos/random-intercepts.html#exvar -#making some random data -sampled_data <- readme_df[sample(nrow(readme_df), 220), ] -expanded_sample_data <- expand_timeseries(sampled_data[1,]) -for (i in 2:nrow(sampled_data)){ -expanded_sample_data <- rbind(expanded_sample_data, expand_timeseries(sampled_data[i,])) -} -windowed_sample_data <- expanded_sample_data |> -filter(week >= (27 - window_num) & week <= (27 + window_num)) |> -mutate(D = ifelse(week > 27, 1, 0)) -windowed_sample_data$scaled_project_age <- scale(windowed_sample_data$age_of_project) -windowed_sample_data$week_offset <- windowed_sample_data$week - 27 -all_actions_sample_data <- windowed_sample_data[which(windowed_sample_data$observation_type == "all"),] -#test model -test_model <- lmer(count ~ D * I(week_offset) + scaled_project_age + (D * I(week_offset)|upstream_vcs_link), data=all_actions_sample_data, REML=FALSE) -summary(test_model) -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=predict(test_model)), show.legend = FALSE) + -theme_bw() -p -# https://www.bristol.ac.uk/cmm/learning/videos/random-intercepts.html#exvar -#making some random data -sampled_data <- readme_df[sample(nrow(readme_df), 220), ] -expanded_sample_data <- expand_timeseries(sampled_data[1,]) -for (i in 2:nrow(sampled_data)){ -expanded_sample_data <- rbind(expanded_sample_data, expand_timeseries(sampled_data[i,])) -} -windowed_sample_data <- expanded_sample_data |> -filter(week >= (27 - window_num) & week <= (27 + window_num)) |> -mutate(D = ifelse(week > 27, 1, 0)) -windowed_sample_data$scaled_project_age <- scale(windowed_sample_data$age_of_project) -windowed_sample_data$week_offset <- windowed_sample_data$week - 27 -all_actions_sample_data <- windowed_sample_data[which(windowed_sample_data$observation_type == "all"),] -#test model -test_model <- lmer(count ~ D * I(week_offset) + scaled_project_age + (D * I(week_offset)|upstream_vcs_link), data=all_actions_sample_data, REML=FALSE) -summary(test_model) -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=predict(test_model)), show.legend = FALSE) + -theme_bw() -p -summary(all_model) -all_residuals <- residuals(all_model) -qqnorm(all_residuals) -mrg_residuals <- residuals(mrg_model) -qqnorm(mrg_residuals) -summary(all_model) -##end of the model testing and plotting section -all_model <- lmer(count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=TRUE) -summary(all_model) -##end of the model testing and plotting section -all_model <- lmer(count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE) -summary(all_model) +library(ggplot2) +data1 <- read_csv('../power_data_111023_mmt.csv',show_col_types = FALSE) +data2 <- read_csv('../inst_all_packages_full_results.csv') +data1 <- read_csv('../kk_final_expanded_data_final.csv',show_col_types = FALSE) +library(readr) +library(ggplot2) +library(tidyverse) +data1 <- read_csv('../kk_final_expanded_data_final.csv',show_col_types = FALSE) # this is the file with the lmer multi-level rddAnalysis library(tidyverse) library(plyr) @@ -272,155 +50,200 @@ windowed_data$week_offset <- windowed_data$week - 27 #separate out the cleaning d all_actions_data <- windowed_data[which(windowed_data$observation_type == "all"),] mrg_actions_data <- windowed_data[which(windowed_data$observation_type == "mrg"),] -#find some EDA to identify which types of models might be the best for this -hist(all_actions_data$count) -#find some EDA to identify which types of models might be the best for this -hist(log1p(all_actions_data$count)) -#find some EDA to identify which types of models might be the best for this -hist(log(all_actions_data$count)) +all_actions_data$log1p_count <- log1p(all_actions_data$count) # 3 rdd in lmer analysis # rdd: https://rpubs.com/phle/r_tutorial_regression_discontinuity_design # lmer: https://www.youtube.com/watch?v=LzAwEKrn2Mc library(lme4) -##end of the model testing and plotting section -all_model <- lmer(log(count) ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE) -all_actions_data$logged_count <- log(all_actions_data$count) -all_actions_data$log1p_count <- log1p(all_actions_data$count) -##end of the model testing and plotting section -all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE) -summary(all_model) -all_residuals <- residuals(all_model) -qqnorm(all_residuals) -##end of the model testing and plotting section -all_model <- lmer(logged_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE) -##end of the model testing and plotting section -all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE) -##end of the model testing and plotting section -all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (week_offset| upstream_vcs_link), data=all_actions_data, REML=FALSE) -##end of the model testing and plotting section -all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE) # https://www.bristol.ac.uk/cmm/learning/videos/random-intercepts.html#exvar -#making some random data -sampled_data <- readme_df[sample(nrow(readme_df), 220), ] -expanded_sample_data <- expand_timeseries(sampled_data[1,]) -for (i in 2:nrow(sampled_data)){ -expanded_sample_data <- rbind(expanded_sample_data, expand_timeseries(sampled_data[i,])) -} -windowed_sample_data <- expanded_sample_data |> -filter(week >= (27 - window_num) & week <= (27 + window_num)) |> -mutate(D = ifelse(week > 27, 1, 0)) -windowed_sample_data$scaled_project_age <- scale(windowed_sample_data$age_of_project) -windowed_sample_data$week_offset <- windowed_sample_data$week - 27 -all_actions_sample_data <- windowed_sample_data[which(windowed_sample_data$observation_type == "all"),] -#test model -test_model <- lmer(log1p_count ~ D * I(week_offset) + scaled_project_age + (D * I(week_offset)|upstream_vcs_link), data=all_actions_sample_data, REML=FALSE) -all_actions_sample_data$log1p_count <- log1p(all_actions_sample_data$count) -#test model -test_model <- lmer(log1p_count ~ D * I(week_offset) + scaled_project_age + (D * I(week_offset)|upstream_vcs_link), data=all_actions_sample_data, REML=FALSE) -summary(test_model) -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=predict(test_model)), show.legend = FALSE) + -theme_bw() -p -#plot results -p <- ggplot(all_actions_sample_data, aes(x=week_offset, y=count, color=upstream_vcs_link), show.legend = FALSE) + -geom_point(size=3, show.legend = FALSE) + -geom_line(aes(y=predict(test_model)), show.legend = FALSE) + -theme_bw() -p -library(merTools) -ICC(outcome="count", group="week", data=all_actions_data) -ICC(outcome="count", group="upstream_vcs_link", data=all_actions_data) -ICC(outcome="count", group="week", data=all_actions_data) -# this is the file with the lmer multi-level rddAnalysis -library(tidyverse) -library(plyr) -# 0 loading the readme data in -try(setwd(dirname(rstudioapi::getActiveDocumentContext()$path))) -readme_df <- read_csv("../final_data/deb_readme_did.csv") -# 1 preprocessing -#colnames(readme_df) <- c("upstream_vcs_link", "event_date", "event_hash", "before_all_ct", "before_mrg_ct", "after_all_ct", "after_mrg_ct", "before_auth_new", "after_commit_new", "after_auth_new", "before_commit_new") -col_order <- c("upstream_vcs_link", "age_of_project", "event_date", "event_hash", "before_all_ct", "after_all_ct", "before_mrg_ct", "after_mrg_ct", "before_auth_new", "after_auth_new", "before_commit_new", "after_commit_new") -readme_df <- readme_df[,col_order] -readme_df$ct_before_all <- str_split(gsub("[][]","", readme_df$before_all_ct), ", ") -readme_df$ct_after_all <- str_split(gsub("[][]","", readme_df$after_all_ct), ", ") -readme_df$ct_before_mrg <- str_split(gsub("[][]","", readme_df$before_mrg_ct), ", ") -readme_df$ct_after_mrg <- str_split(gsub("[][]","", readme_df$after_mrg_ct), ", ") -drop <- c("before_all_ct", "before_mrg_ct", "after_all_ct", "after_mrg_ct") -readme_df = readme_df[,!(names(readme_df) %in% drop)] -# 2 some expansion needs to happens for each project -expand_timeseries <- function(project_row) { -longer <- project_row |> -pivot_longer(cols = starts_with("ct"), -names_to = "window", -values_to = "count") |> -unnest(count) -longer$observation_type <- gsub("^.*_", "", longer$window) -longer <- ddply(longer, "observation_type", transform, week=seq(from=0, by=1, length.out=length(observation_type))) -longer$count <- as.numeric(longer$count) -#longer <- longer[which(longer$observation_type == "all"),] -return(longer) -} -expanded_data <- expand_timeseries(readme_df[1,]) -for (i in 2:nrow(readme_df)){ -expanded_data <- rbind(expanded_data, expand_timeseries(readme_df[i,])) -} -#filter out the windows of time that we're looking at -window_num <- 8 -windowed_data <- expanded_data |> -filter(week >= (27 - window_num) & week <= (27 + window_num)) |> -mutate(D = ifelse(week > 27, 1, 0)) -#scale the age numbers -windowed_data$scaled_project_age <- scale(windowed_data$age_of_project) -windowed_data$week_offset <- windowed_data$week - 27 -#separate out the cleaning d -all_actions_data <- windowed_data[which(windowed_data$observation_type == "all"),] -mrg_actions_data <- windowed_data[which(windowed_data$observation_type == "mrg"),] -all_actions_data$logged_count <- log(all_actions_data$count) -all_actions_data$log1p_count <- log1p(all_actions_data$count) -# 3 rdd in lmer analysis -# rdd: https://rpubs.com/phle/r_tutorial_regression_discontinuity_design -# lmer: https://www.youtube.com/watch?v=LzAwEKrn2Mc -library(lme4) -##end of the model testing and plotting section -all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE) -summary(all_model) -( -##end of the model testing and plotting section -all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE, control = lmerControl(optimizer ="Nelder_Mead")) -##end of the model testing and plotting section -all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE, control = lmerControl(optimizer ="Nelder_Mead")) -all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE, control = lmerControl(optimizer="optimx", -optCtrl=list(method="nlminb"))) +library(optimx) +library(lattice) all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE, control = lmerControl( optimizer ='optimx', optCtrl=list(method='L-BFGS-B'))) -summary(all_model) -all0_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE) -summary(all0_model) -all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE, control = lmerControl( -optimizer ='optimx', optCtrl=list(method='L-BFGS-B'))) -all_model.ranef <- ranef(all_model) -str(all_model.ranef) -head(all_model.ranef) -all_actions_data$D_effect_quart <- ntile(all_model.ranef$D, 4) -head(all_model.ranef) -all_model.ranef <- random.effects(all_model) -head(as.data.frame(all_model.ranef)) -head(all_model_ranef) -all_model_ranef <- as.data.frame(ranef(all_model)) -head(all_model_ranef) -d_effect_ranef_all <- subset(all_model_ranef, term="D") -d_effect_ranef_all <- all_model_ranef[all_model_ranef$term="D",] +#identifying the quartiles of effect for D +all_model_ranef <- as.data.frame(ranef(all_model, condVar=TRUE)) +View(all_model_ranef) d_effect_ranef_all <- all_model_ranef[all_model_ranef$term=="D",] +#identifying the quartiles of effect for D +all_model_ranef <- ranef(all_model, condVar=TRUE) +d_effect_ranef_all <- all_model_ranef[all_model_ranef$term=="D",] +dotplot(all_model_ranef) +d_effect_ranef_all <- all_model_ranef['upstream_vcs_link']['D'] +View(all_model_ranef) +d_effect_ranef_all <- all_model_ranef[upstream_vcs_link,2] +d_effect_ranef_all <- all_model_ranef['upstream_vcs_link',2] +d_effect_ranef_all <- all_model_ranef$upstream_vcs_link View(d_effect_ranef_all) -d_effect_ranef_all$quartile <- ntile(d_effect_ranef_all$condval, 4) -View(d_effect_ranef_all) -# mrg behavior for this -mrg_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE, control = lmerControl( +dotplot(all_model_ranef)[["D"]] +dotplot(all_model_ranef)[["upstream_vcs_link"]] +dotplot(all_model_ranef)[["upstream_vcs_link"]["D"]] +dotplot(all_model_ranef)$D +View(all_model_ranef) +for (j in 1:nschool) { +jj <- order(all_model_ranef)[j] +lines (x=c(j,j),y=c(ranef.lower[jj],ranef.upper[jj])) +} +for (j in 1:upstream_vcs_link) { +jj <- order(all_model_ranef)[j] +lines (x=c(j,j),y=c(ranef.lower[jj],ranef.upper[jj])) +} +View(all_model_ranef) +df_ranefs <- as.data.frame(all_model_ranef) +View(df_ranefs) +#identifying the quartiles of effect for D +all_model_ranef <- ranef(all_model, condVar=TRUE)$upstream_vcs_link[[2]] +#identifying the quartiles of effect for D +all_model_ranef <- ranef(all_model, condVar=TRUE)$upstream_vcs_link +dotplot(all_model_ranef) +dotplot(all_model_ranef) +# this is the file with the lmer multi-level rddAnalysis +library(tidyverse) +library(plyr) +# 0 loading the readme data in +try(setwd(dirname(rstudioapi::getActiveDocumentContext()$path))) +readme_df <- read_csv("../final_data/deb_readme_did.csv") +# 1 preprocessing +#colnames(readme_df) <- c("upstream_vcs_link", "event_date", "event_hash", "before_all_ct", "before_mrg_ct", "after_all_ct", "after_mrg_ct", "before_auth_new", "after_commit_new", "after_auth_new", "before_commit_new") +col_order <- c("upstream_vcs_link", "age_of_project", "event_date", "event_hash", "before_all_ct", "after_all_ct", "before_mrg_ct", "after_mrg_ct", "before_auth_new", "after_auth_new", "before_commit_new", "after_commit_new") +readme_df <- readme_df[,col_order] +readme_df$ct_before_all <- str_split(gsub("[][]","", readme_df$before_all_ct), ", ") +readme_df$ct_after_all <- str_split(gsub("[][]","", readme_df$after_all_ct), ", ") +readme_df$ct_before_mrg <- str_split(gsub("[][]","", readme_df$before_mrg_ct), ", ") +readme_df$ct_after_mrg <- str_split(gsub("[][]","", readme_df$after_mrg_ct), ", ") +drop <- c("before_all_ct", "before_mrg_ct", "after_all_ct", "after_mrg_ct") +readme_df = readme_df[,!(names(readme_df) %in% drop)] +# 2 some expansion needs to happens for each project +expand_timeseries <- function(project_row) { +longer <- project_row |> +pivot_longer(cols = starts_with("ct"), +names_to = "window", +values_to = "count") |> +unnest(count) +longer$observation_type <- gsub("^.*_", "", longer$window) +longer <- ddply(longer, "observation_type", transform, week=seq(from=0, by=1, length.out=length(observation_type))) +longer$count <- as.numeric(longer$count) +#longer <- longer[which(longer$observation_type == "all"),] +return(longer) +} +expanded_data <- expand_timeseries(readme_df[1,]) +for (i in 2:nrow(readme_df)){ +expanded_data <- rbind(expanded_data, expand_timeseries(readme_df[i,])) +} +#filter out the windows of time that we're looking at +window_num <- 8 +windowed_data <- expanded_data |> +filter(week >= (27 - window_num) & week <= (27 + window_num)) |> +mutate(D = ifelse(week > 27, 1, 0)) +#scale the age numbers +windowed_data$scaled_project_age <- scale(windowed_data$age_of_project) +windowed_data$week_offset <- windowed_data$week - 27 +#separate out the cleaning d +all_actions_data <- windowed_data[which(windowed_data$observation_type == "all"),] +mrg_actions_data <- windowed_data[which(windowed_data$observation_type == "mrg"),] +all_actions_data$log1p_count <- log1p(all_actions_data$count) +# 3 rdd in lmer analysis +# rdd: https://rpubs.com/phle/r_tutorial_regression_discontinuity_design +# lmer: https://www.youtube.com/watch?v=LzAwEKrn2Mc +library(lme4) +# https://www.bristol.ac.uk/cmm/learning/videos/random-intercepts.html#exvar +library(optimx) +library(lattice) +all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE, control = lmerControl( optimizer ='optimx', optCtrl=list(method='L-BFGS-B'))) -summary(mrg_model) +#identifying the quartiles of effect for D +all_model_ranef <- ranef(all_model, condVar=TRUE) +dotplot(all_model_ranef)$g[1] +re <- ranef(all_model,postVar = TRUE) +re$g$'(Intercept)' <- NULL +re$g$'D:I(week_offset)' <- NULL +re <- ranef(all_model, condVar=TRUE) +re$g$'(Intercept)' <- NULL +re$g$'D:I(week_offset)' <- NULL +dotplot(re) +dotplot(re) +re <- ranef(all_model, condVar=TRUE) +re$upstream_vcs_link$'(Intercept)' <- NULL +re$upstream_vcs_link$'D:I(week_offset)' <- NULL +View(re) +re$upstream_vcs_link$'I(week_offset)' <- NULL +dotplot(re) +View(re) +View(all_model_ranef) +dotplot(all_model_ranef) +#identifying the quartiles of effect for D +all_model_ranef <- ranef(all_model, condVar=TRUE) +dotplot(all_model_ranef) +#identifying the quartiles of effect for D +all_model_ranef <- ranef(all_model, condVar=TRUE) +dotplot(all_model_ranef) +re <- ranef(all_model, condVar=TRUE) +re$upstream_vcs_link$'(Intercept)' <- NULL +re$upstream_vcs_link$'D:I(week_offset)' <- NULL +re$upstream_vcs_link$'I(week_offset)' <- NULL +dotplot(re) +dotplot(all_model_ranef) +# this is the file with the lmer multi-level rddAnalysis +library(tidyverse) +library(plyr) +# 0 loading the readme data in +try(setwd(dirname(rstudioapi::getActiveDocumentContext()$path))) +readme_df <- read_csv("../final_data/deb_readme_did.csv") +# 1 preprocessing +#colnames(readme_df) <- c("upstream_vcs_link", "event_date", "event_hash", "before_all_ct", "before_mrg_ct", "after_all_ct", "after_mrg_ct", "before_auth_new", "after_commit_new", "after_auth_new", "before_commit_new") +col_order <- c("upstream_vcs_link", "age_of_project", "event_date", "event_hash", "before_all_ct", "after_all_ct", "before_mrg_ct", "after_mrg_ct", "before_auth_new", "after_auth_new", "before_commit_new", "after_commit_new") +readme_df <- readme_df[,col_order] +readme_df$ct_before_all <- str_split(gsub("[][]","", readme_df$before_all_ct), ", ") +readme_df$ct_after_all <- str_split(gsub("[][]","", readme_df$after_all_ct), ", ") +readme_df$ct_before_mrg <- str_split(gsub("[][]","", readme_df$before_mrg_ct), ", ") +readme_df$ct_after_mrg <- str_split(gsub("[][]","", readme_df$after_mrg_ct), ", ") +drop <- c("before_all_ct", "before_mrg_ct", "after_all_ct", "after_mrg_ct") +readme_df = readme_df[,!(names(readme_df) %in% drop)] +# 2 some expansion needs to happens for each project +expand_timeseries <- function(project_row) { +longer <- project_row |> +pivot_longer(cols = starts_with("ct"), +names_to = "window", +values_to = "count") |> +unnest(count) +longer$observation_type <- gsub("^.*_", "", longer$window) +longer <- ddply(longer, "observation_type", transform, week=seq(from=0, by=1, length.out=length(observation_type))) +longer$count <- as.numeric(longer$count) +#longer <- longer[which(longer$observation_type == "all"),] +return(longer) +} +expanded_data <- expand_timeseries(readme_df[1,]) +for (i in 2:nrow(readme_df)){ +expanded_data <- rbind(expanded_data, expand_timeseries(readme_df[i,])) +} +#filter out the windows of time that we're looking at +window_num <- 8 +windowed_data <- expanded_data |> +filter(week >= (27 - window_num) & week <= (27 + window_num)) |> +mutate(D = ifelse(week > 27, 1, 0)) +#scale the age numbers +windowed_data$scaled_project_age <- scale(windowed_data$age_of_project) +windowed_data$week_offset <- windowed_data$week - 27 +#separate out the cleaning d +all_actions_data <- windowed_data[which(windowed_data$observation_type == "all"),] +mrg_actions_data <- windowed_data[which(windowed_data$observation_type == "mrg"),] +all_actions_data$log1p_count <- log1p(all_actions_data$count) +# 3 rdd in lmer analysis +# rdd: https://rpubs.com/phle/r_tutorial_regression_discontinuity_design +# lmer: https://www.youtube.com/watch?v=LzAwEKrn2Mc +library(lme4) +# https://www.bristol.ac.uk/cmm/learning/videos/random-intercepts.html#exvar +library(optimx) +library(lattice) +all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE, control = lmerControl( +optimizer ='optimx', optCtrl=list(method='L-BFGS-B'))) +summary(all_model) +#identifying the quartiles of effect for D +all_model_ranef <- ranef(all_model, condVar=TRUE) +dotplot(all_model_ranef) +df_ranefs <- as.data.frame(all_model_ranef) +View(df_ranefs) +D_df_ranef <- df_ranefs[term == "D"] +D_df_ranef <- df_ranefs[df_ranefs$term == "D"] library(tidyverse) library(plyr) #get the contrib data instead @@ -464,49 +287,226 @@ windowed_data$week_offset <- windowed_data$week - 27 all_actions_data <- windowed_data[which(windowed_data$observation_type == "all"),] mrg_actions_data <- windowed_data[which(windowed_data$observation_type == "mrg"),] #EDA? +hist(log(all_actions_data$count)) +all_actions_data$logged_count <- log(all_actions_data$count) +all_actions_data$log1p_count <- log1p(all_actions_data$count) +# now for merge +mrg_actions_data$logged_count <- log(mrg_actions_data$count) +mrg_actions_data$log1p_count <- log1p(mrg_actions_data$count) #TKTK --------------------- #imports for models library(lme4) library(optimx) -#models -- TKTK need to be fixed -all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE, control = lmerControl( -optimizer ='optimx', optCtrl=list(method='L-BFGS-B'))) -#EDA? -hist(log(all_actions_data$count)) -all_actions_data$logged_count <- log(all_actions_data$count) -all_actions_data$log1p_count <- log1p(all_actions_data$count) -#models -- TKTK need to be fixed -all_model <- lmer(logged_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE, control = lmerControl( -optimizer ='optimx', optCtrl=list(method='L-BFGS-B'))) -#models -- TKTK need to be fixed -all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE, control = lmerControl( -optimizer ='optimx', optCtrl=list(method='L-BFGS-B'))) -summary(all_model) +library(lattice) #models -- TKTK need to be fixed all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (week_offset| upstream_vcs_link), data=all_actions_data, REML=FALSE, control = lmerControl( optimizer ='optimx', optCtrl=list(method='L-BFGS-B'))) summary(all_model) -#models -- TKTK need to be fixed -all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE) -week_offset -#models -- TKTK need to be fixed -all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (week_offset| upstream_vcs_link), data=all_actions_data, REML=FALSE, control = lmerControl( -optimizer ='optimx', optCtrl=list(method='L-BFGS-B'))) -summary(all_model) -# mrg behavior for this -mrg_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE, control = lmerControl( -optimizer ='optimx', optCtrl=list(method='L-BFGS-B'))) -# mrg behavior for this -mrg_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (week_offset | upstream_vcs_link), data=all_actions_data, REML=FALSE, control = lmerControl( -optimizer ='optimx', optCtrl=list(method='L-BFGS-B'))) -summary(mrg_model) -# mrg behavior for this -mrg_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (week_offset | upstream_vcs_link), data=mrg_actions_data, REML=FALSE, control = lmerControl( -optimizer ='optimx', optCtrl=list(method='L-BFGS-B'))) -# now for merge -mrg_actions_data$logged_count <- log(mrg_actions_data$count) -mrg_actions_data$log1p_count <- log1p(mrg_actions_data$count) +#identifying the quartiles of effect for D +all_model_ranef <- as.data.frame(ranef(all_model)) +View(all_model_ranef) # mrg behavior for this mrg_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (week_offset | upstream_vcs_link), data=mrg_actions_data, REML=FALSE, control = lmerControl( optimizer ='optimx', optCtrl=list(method='L-BFGS-B'))) summary(mrg_model) +#identifying the quartiles of effect for D +mrg_model_ranef <- ranef(mrg_model) +View(mrg_model_ranef) +dotplot(mrg_model_ranef) +#load in data +contrib_df <- read_csv("../final_data/deb_contrib_pop_change.csv") +readme_df <- read_csv("../final_data/deb_readme_pop_change.csv") +View(readme_df) +#some expansion needs to happens for each project +expand_timeseries <- function(project_row) { +longer <- project_row |> +pivot_longer(cols = ends_with("new"), +names_to = "window", +values_to = "count") |> +unnest(count) +#longer$observation_type <- gsub("^.*_", "", longer$window) +#longer <- ddply(longer, "observation_type", transform, week=seq(from=0, by=1, length.out=length(observation_type))) +#longer$count <- as.numeric(longer$count) +#longer <- longer[which(longer$observation_type == "all"),] +return(longer) +} +expanded_data <- expand_timeseries(readme_df[1,]) +View(expand_timeseries) +View(expanded_data) +longer <- project_row |> +pivot_longer(cols = ends_with("new"), +names_to = "window", +values_to = "count") |> +unnest(count) |> +mutate(after_doc = str_detect(window, "after")) +#some expansion needs to happens for each project +expand_timeseries <- function(project_row) { +longer <- project_row |> +pivot_longer(cols = ends_with("new"), +names_to = "window", +values_to = "count") |> +unnest(count) |> +mutate(after_doc = str_detect(window, "after")) +return(longer) +} +expanded_data <- expand_timeseries(readme_df[1,]) +longer <- project_row |> +pivot_longer(cols = ends_with("new"), +names_to = "window", +values_to = "count") |> +unnest(count) |> +mutate(after_doc = as.numeric(str_detect(window, "after"))) +return(longer) +#some expansion needs to happens for each project +expand_timeseries <- function(project_row) { +longer <- project_row |> +pivot_longer(cols = ends_with("new"), +names_to = "window", +values_to = "count") |> +unnest(count) |> +mutate(after_doc = as.numeric(str_detect(window, "after"))) +return(longer) +} +expanded_data <- expand_timeseries(readme_df[1,]) +#some expansion needs to happens for each project +expand_timeseries <- function(project_row) { +longer <- project_row |> +pivot_longer(cols = ends_with("new"), +names_to = "window", +values_to = "count") |> +unnest(count) |> +mutate(after_doc = as.numeric(str_detect(window, "after"))) |> +mutate(is_collab = as.numeric(str_detect(window, "collab"))) +return(longer) +} +expanded_data <- expand_timeseries(readme_df[1,]) +for (i in 2:nrow(readme_df)){ +expanded_readme_data <- rbind(expanded_readme_data, expand_timeseries(readme_df[i,])) +} +expanded_readme_data <- expand_timeseries(readme_df[1,]) +for (i in 2:nrow(readme_df)){ +expanded_readme_data <- rbind(expanded_readme_data, expand_timeseries(readme_df[i,])) +} +expanded_contrib_data <- expand_timeseries(contrib_df[1,]) +for (i in 2:nrow(contrib_df)){ +expanded_contrib_data <- rbind(expanded_contrib_data, expand_timeseries(contrib_df[i,])) +} +View(expanded_contrib_data) +readme_model <- lmer(count ~ after_doc + (1| upstream_vcs_link), data=expanded_readme_data, REML=FALSE) +summary(readme_model) +readme_model <- lmer(count ~ after_doc + ( after_doc | upstream_vcs_link), data=expanded_readme_data, REML=FALSE) +summary(readme_model) +contrib_model <- lmer(count ~ after_doc + ( after_doc | upstream_vcs_link), data=expanded_contrib_data, REML=FALSE) +summary(contrib_model) +collab_readme_model <- lmer(count ~ after_doc + ( after_doc | upstream_vcs_link), data=collab_pop_readme, REML=FALSE) +#breaking out the types of population counts +collab_pop_readme <- expanded_readme_data[which(expanded_readme_data$is_collab == 1),] +contrib_pop_readme <- expanded_readme_data[which(expanded_readme_data$is_collab == 0),] +collab_pop_contrib <- expanded_contrib_data[which(expanded_contrib_data$is_collab == 1),] +contrib_pop_contrib <- expanded_contrib_data[which(expanded_contrib_data$is_collab == 0),] +collab_readme_model <- lmer(count ~ after_doc + ( after_doc | upstream_vcs_link), data=collab_pop_readme, REML=FALSE) +collab_readme_model <- lmer(count ~ after_doc + (1| upstream_vcs_link), data=collab_pop_readme, REML=FALSE) +summary(collab_readme_model) +contrib_readme_model <- lmer(count ~ after_doc + ( 1| upstream_vcs_link), data=contrib_pop_readme, REML=FALSE) +summary(contrib_readme_model) +collab_readme_model <- lmer(count ~ after_doc + (after_doc| upstream_vcs_link), data=collab_pop_readme, REML=FALSE) +collab_readme_model <- lmer(count ~ after_doc + (1| upstream_vcs_link), data=collab_pop_readme, REML=FALSE) +summary(collab_readme_model) +contrib_readme_model <- lmer(count ~ after_doc + ( 1| upstream_vcs_link), data=contrib_pop_readme, REML=FALSE) +summary(contrib_readme_model) +collab_contrib_model <- lmer(count ~ after_doc + ( 1 | upstream_vcs_link), data=collab_pop_contrib, REML=FALSE) +summary(collab_contrib_model) +contrib_contrib_model <- lmer(count ~ after_doc + ( 1 | upstream_vcs_link), data=contrib_pop_contrib, REML=FALSE) +summary(contrib_contrib_model) +expanded_readme_data |> +ggplot(aes(x = after_doc, y = count, col = is_collab)) + +geom_point() +ggplot(aes(x = after_doc, y = count, col = as.factor(is_collab)) + +expanded_readme_data |> +ggplot(aes(x = after_doc, y = count, col = as.factor(is_collab))) + +geom_point() +expanded_readme_data |> +expanded_readme_data |> +ggplot(aes(x = after_doc, y = count, col = as.factor(is_collab))) + +geom_point() +expanded_readme_data |> +ggplot(aes(x = as.factor(after_doc), y = count, col = as.factor(is_collab))) + +geom_point() +expanded_readme_data |> +ggplot(aes(x = as.factor(after_doc), y = scale(count), col = as.factor(is_collab))) + +geom_point() +expanded_readme_data$log1pcount <- log1p(expanded_readme_data$count) +expanded_contrib_data$log1pcount <- log1p(expanded_contrib_data$count) +expanded_readme_data |> +ggplot(aes(x = as.factor(after_doc), y = log1pcount, col = as.factor(is_collab))) + +geom_point() +expanded_readme_data |> +ggplot(aes(x = as.factor(after_doc), y = log1pcount, col = as.factor(is_collab))) + +geom_point() + +geom_smooth(method = 'lm', se = F) +expanded_readme_data |> +ggplot(aes(x = after_doc, y = log1pcount, col = as.factor(is_collab))) + +geom_point() + +geom_smooth(method = 'lm', se = F) +expanded_contrib_data |> +ggplot(aes(x = after_doc, y = log1pcount, col = as.factor(is_collab))) + +geom_point() + +geom_smooth(method = 'lm', se = F) +expanded_readme_data |> +ggplot(aes(x = after_doc, y = log1pcount, col = as.factor(is_collab))) + +geom_point() + +geom_smooth(method = 'lm', se = F) + geom_jitter() +expanded_readme_data |> +ggplot(aes(x = after_doc, y = count, col = as.factor(is_collab))) + +geom_point() + +geom_smooth(method = 'lm', se = F) + geom_jitter() +expanded_readme_data$logcount <- log(expanded_readme_data$count) +expanded_contrib_data$logcount <- log(expanded_contrib_data$count) +expanded_readme_data |> +ggplot(aes(x = after_doc, y = logcount, col = as.factor(is_collab))) + +geom_point() + +geom_smooth(method = 'lm', se = F) + geom_jitter() +expanded_contrib_data |> +ggplot(aes(x = after_doc, y = log1pcount, col = as.factor(is_collab))) + +geom_point() + +geom_smooth(method = 'lm', se = F) +expanded_contrib_data |> +ggplot(aes(x = after_doc, y = log1pcount, col = as.factor(is_collab))) + +geom_point() + +geom_smooth(method = 'lm', se = F) + geom_jitter() +expanded_contrib_data |> +ggplot(aes(x = after_doc, y = count, col = as.factor(is_collab))) + +geom_point() + +geom_smooth(method = 'lm', se = F) + geom_jitter() +expanded_readme_data |> +ggplot(aes(x = after_doc, y = logcount, col = as.factor(is_collab))) + +geom_point() + +geom_smooth(method = 'lm', se = F) + geom_jitter() +collab_readme_model <- lmer(logcount ~ after_doc + (1| upstream_vcs_link), data=collab_pop_readme, REML=FALSE) +#breaking out the types of population counts +collab_pop_readme <- expanded_readme_data[which(expanded_readme_data$is_collab == 1),] +contrib_pop_readme <- expanded_readme_data[which(expanded_readme_data$is_collab == 0),] +collab_pop_contrib <- expanded_contrib_data[which(expanded_contrib_data$is_collab == 1),] +contrib_pop_contrib <- expanded_contrib_data[which(expanded_contrib_data$is_collab == 0),] +collab_readme_model <- lmer(logcount ~ after_doc + (1| upstream_vcs_link), data=collab_pop_readme, REML=FALSE) +collab_readme_model <- lmer(log1pcount ~ after_doc + (1| upstream_vcs_link), data=collab_pop_readme, REML=FALSE) +summary(collab_readme_model) +contrib_readme_model <- lmer(log1pcount ~ after_doc + ( 1| upstream_vcs_link), data=contrib_pop_readme, REML=FALSE) +summary(contrib_readme_model) +collab_contrib_model <- lmer(log1pcount ~ after_doc + ( 1 | upstream_vcs_link), data=collab_pop_contrib, REML=FALSE) +summary(collab_contrib_model) +contrib_contrib_model <- lmer(log1pcount ~ after_doc + ( 1 | upstream_vcs_link), data=contrib_pop_contrib, REML=FALSE) +summary(contrib_contrib_model) +library(ggplot2) +expanded_readme_data |> +ggplot(aes(x = after_doc, y = logcount, col = as.factor(is_collab))) + +geom_point() + +geom_smooth(method = 'lm', se = F) + geom_jitter() +expanded_readme_data |> +ggplot(aes(x = after_doc, y = log1pcount, col = as.factor(is_collab))) + +geom_point() + +geom_smooth(method = 'lm', se = F) + geom_jitter() +expanded_contrib_data |> +ggplot(aes(x = after_doc, y = count, col = as.factor(is_collab))) + +geom_point() + +geom_smooth(method = 'lm', se = F) + geom_jitter() diff --git a/R/contribRDDAnalysis.R b/R/contribRDDAnalysis.R index 90d970c..24ebdc6 100644 --- a/R/contribRDDAnalysis.R +++ b/R/contribRDDAnalysis.R @@ -51,12 +51,13 @@ mrg_actions_data$log1p_count <- log1p(mrg_actions_data$count) #imports for models library(lme4) library(optimx) +library(lattice) #models -- TKTK need to be fixed all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (week_offset| upstream_vcs_link), data=all_actions_data, REML=FALSE, control = lmerControl( optimizer ='optimx', optCtrl=list(method='L-BFGS-B'))) summary(all_model) #identifying the quartiles of effect for D -all_model_ranef <- as.data.frame(ranef(all_model)) +all_model_ranef <- ranef(all_model) d_effect_ranef_all <- all_model_ranef[all_model_ranef$term=="D",] d_effect_ranef_all$quartile <- ntile(d_effect_ranef_all$condval, 4) #model residuals @@ -67,12 +68,10 @@ mrg_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (week_o optimizer ='optimx', optCtrl=list(method='L-BFGS-B'))) summary(mrg_model) #identifying the quartiles of effect for D -mrg_model_ranef <- as.data.frame(ranef(mrg_model)) +mrg_model_ranef <- ranef(mrg_model) +dotplot(mrg_model_ranef) d_effect_ranef_mrg <- mrg_model_ranef[mrg_model_ranef$term=="D",] d_effect_ranef_mrg$quartile <- ntile(d_effect_ranef_mrg$condval, 4) #merge model residuals mrg_residuals <- residuals(mrg_model) -qqnorm(mrg_residuals) - - - +qqnorm(mrg_residuals) \ No newline at end of file diff --git a/R/popRDDAnalyssis.R b/R/popRDDAnalyssis.R new file mode 100644 index 0000000..951fd78 --- /dev/null +++ b/R/popRDDAnalyssis.R @@ -0,0 +1,55 @@ +library(tidyverse) +library(plyr) +library(stringr) +try(setwd(dirname(rstudioapi::getActiveDocumentContext()$path))) +#load in data +contrib_df <- read_csv("../final_data/deb_contrib_pop_change.csv") +readme_df <- read_csv("../final_data/deb_readme_pop_change.csv") +#some expansion needs to happens for each project +expand_timeseries <- function(project_row) { + longer <- project_row |> + pivot_longer(cols = ends_with("new"), + names_to = "window", + values_to = "count") |> + unnest(count) |> + mutate(after_doc = as.numeric(str_detect(window, "after"))) |> + mutate(is_collab = as.numeric(str_detect(window, "collab"))) + return(longer) +} +expanded_readme_data <- expand_timeseries(readme_df[1,]) +for (i in 2:nrow(readme_df)){ + expanded_readme_data <- rbind(expanded_readme_data, expand_timeseries(readme_df[i,])) +} +expanded_contrib_data <- expand_timeseries(contrib_df[1,]) +for (i in 2:nrow(contrib_df)){ + expanded_contrib_data <- rbind(expanded_contrib_data, expand_timeseries(contrib_df[i,])) +} +expanded_readme_data$log1pcount <- log1p(expanded_readme_data$count) +expanded_contrib_data$log1pcount <- log1p(expanded_contrib_data$count) +expanded_readme_data$logcount <- log(expanded_readme_data$count) +expanded_contrib_data$logcount <- log(expanded_contrib_data$count) +#breaking out the types of population counts +collab_pop_readme <- expanded_readme_data[which(expanded_readme_data$is_collab == 1),] +contrib_pop_readme <- expanded_readme_data[which(expanded_readme_data$is_collab == 0),] +collab_pop_contrib <- expanded_contrib_data[which(expanded_contrib_data$is_collab == 1),] +contrib_pop_contrib <- expanded_contrib_data[which(expanded_contrib_data$is_collab == 0),] +#import models +library(lme4) +library(optimx) +collab_readme_model <- lmer(log1pcount ~ after_doc + (1| upstream_vcs_link), data=collab_pop_readme, REML=FALSE) +summary(collab_readme_model) +contrib_readme_model <- lmer(log1pcount ~ after_doc + ( 1| upstream_vcs_link), data=contrib_pop_readme, REML=FALSE) +summary(contrib_readme_model) +collab_contrib_model <- lmer(log1pcount ~ after_doc + ( 1 | upstream_vcs_link), data=collab_pop_contrib, REML=FALSE) +summary(collab_contrib_model) +contrib_contrib_model <- lmer(log1pcount ~ after_doc + ( 1 | upstream_vcs_link), data=contrib_pop_contrib, REML=FALSE) +summary(contrib_contrib_model) + +library(ggplot2) +expanded_readme_data |> + ggplot(aes(x = after_doc, y = log1pcount, col = as.factor(is_collab))) + + geom_point() + geom_jitter() + +expanded_contrib_data |> + ggplot(aes(x = after_doc, y = count, col = as.factor(is_collab))) + + geom_point() + geom_jitter() diff --git a/R/readmeRDDAnalysis.R b/R/readmeRDDAnalysis.R index 6cf75a1..6640380 100644 --- a/R/readmeRDDAnalysis.R +++ b/R/readmeRDDAnalysis.R @@ -58,13 +58,17 @@ all_actions_data$log1p_count <- log1p(all_actions_data$count) library(lme4) # https://www.bristol.ac.uk/cmm/learning/videos/random-intercepts.html#exvar library(optimx) +library(lattice) all_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I(week_offset)| upstream_vcs_link), data=all_actions_data, REML=FALSE, control = lmerControl( optimizer ='optimx', optCtrl=list(method='L-BFGS-B'))) summary(all_model) #identifying the quartiles of effect for D -all_model_ranef <- as.data.frame(ranef(all_model)) -d_effect_ranef_all <- all_model_ranef[all_model_ranef$term=="D",] -d_effect_ranef_all$quartile <- ntile(d_effect_ranef_all$condval, 4) +all_model_ranef <- ranef(all_model, condVar=TRUE) +dotplot(all_model_ranef) +df_ranefs <- as.data.frame(all_model_ranef) +#D_df_ranef <- df_ranefs[df_ranefs$term == "D"] +#d_effect_ranef_all <- all_model_ranef$upstream_vcs_link +#d_effect_ranef_all$quartile <- ntile(d_effect_ranef_all$condval, 4) #model residuals all_residuals <- residuals(all_model) qqnorm(all_residuals) @@ -73,12 +77,11 @@ mrg_model <- lmer(log1p_count ~ D * I(week_offset)+ scaled_project_age + (D * I( optimizer ='optimx', optCtrl=list(method='L-BFGS-B'))) summary(mrg_model) #identifying the quartiles of effect for D -mrg_model_ranef <- as.data.frame(ranef(mrg_model)) +mrg_model_ranef <- ranef(mrg_model, condVar=TRUE) +dotplot(mrg_model_ranef) d_effect_ranef_mrg <- mrg_model_ranef[mrg_model_ranef$term=="D",] d_effect_ranef_mrg$quartile <- ntile(d_effect_ranef_mrg$condval, 4) #merge model residuals mrg_residuals <- residuals(mrg_model) qqnorm(mrg_residuals) -# Performance: - - +# Performance: \ No newline at end of file diff --git a/042724_draft_deb_contrib_person.csv b/final_data/deb_contrib_pop_change.csv similarity index 100% rename from 042724_draft_deb_contrib_person.csv rename to final_data/deb_contrib_pop_change.csv diff --git a/42524_new_readme.csv b/final_data/deb_readme_pop_change.csv similarity index 100% rename from 42524_new_readme.csv rename to final_data/deb_readme_pop_change.csv