In OpenSDP/OpenSDPsynthR: Package to generate synthetic data for the OpenSDP Project

knitr::opts_chunk$set(echo = TRUE)
library(OpenSDPsynthR)
simouts <- simpop(nstu = 1250L, seed = 488234, 
                  control = sim_control(nschls = 9L, 
                                        minyear=1997, maxyear=2013))

Identify

names(simouts$assessment)
names(simouts$demog_master)
names(simouts$stu_year)

Clean

TBD

Connect

analyze_table <- left_join(simouts$demog_master, simouts$assessment, 
                           by = "sid")
analyze_table <- left_join(analyze_table, simouts$stu_year, 
                           by = c("sid", "year"))

analyze_table %>% group_by(sid) %>% 
  summarize(nobs = n()) %>% select(-sid) %>% summary

analyze_table <- analyze_table %>% filter(grade.x %in% 
                                            c("3", "4", "5", "6", "7"))

Analyze

# simple

simp_mod <- lm(math_ss ~ Race + grade.x + frpl + ell + iep + gifted, 
               data = analyze_table)

summary(simp_mod)

# school effect

sch_mod <- lm(math_ss ~ Race + grade.x + frpl + ell + iep + gifted + 
                schid.x, 
               data = analyze_table)

summary(sch_mod)

Report

library(broom)
library(ggplot2)

plot_mod <- tidy(sch_mod)
plot_mod$model <- "school"
tmp <- tidy(simp_mod)
tmp$model <- "simple"
plot_mod <- bind_rows(plot_mod, tmp); rm(tmp)


ggplot(plot_mod[plot_mod$term %in% c("frpl1", "ell1", "iep1"), ], 
       aes(x = term, y = estimate, ymin = estimate-std.error, 
           ymax = estimate+std.error, group = model, color = model)) + 
  geom_linerange(position = position_dodge(width = 1), size = 2) + 
  theme_bw() + geom_hline(yintercept = 0, color = I("red")) + 
  theme(legend.position = "bottom")

ggplot(plot_mod[plot_mod$term %in% c("schid.x1", "schid.x2", 
                                     "schid.x3", "schid.x4", 
                                     "schid.x5", "schid.x6", 
                                     "schid.x7", "schid.x8", 
                                     "schid.x9"), ], 
       aes(x = term, y = estimate, ymin = estimate-std.error, 
           ymax = estimate+std.error)) + 
  geom_linerange(size = 2) + 
  theme_bw() + geom_hline(yintercept = 0, color = I("red")) + 
  theme(legend.position = "bottom")

OpenSDP/OpenSDPsynthR documentation built on June 20, 2020, 6:18 a.m.