R/summary_statistics.R

In mabafaba/hypergrammaR:

percent_with_confints <- function(dependent.var,
                                  independent.var,
                                  design,
                                  na.rm = TRUE){



  formula_string<-paste0("~",independent.var, "+",dependent.var )
  f.table <- svytable(formula(formula_string), design)

  # if(design$variables %>%
  #    split.data.frame(design$variables[[independent.var]]) %>%
  #    lapply(nrow) %>%
  #    unlist %>%
  #    (function(x){x<2}) %>%
  #    any){
  #       warning("independent var must have at least two unique values");return(NULL)
  #     }

  formula_err <- paste0("~", dependent.var, sep = "")
  by <- paste0(" ~",independent.var , sep = "")
  summary.result.svyby <- svyby(formula(formula_err), formula(by), design, na.rm = T, svymean)
  # fix weird column names:
  colnames(summary.result.svyby)<-colnames(summary.result.svyby) %>% gsub(paste0("^se.",dependent.var),"",.) %>% gsub(paste0("^",dependent.var),"",.)




  if(length(summary.result.svyby)==3 & length(summary.result.svyby[[2]])==2){return(list(ERROR="error calculating convidendce intervals. illformated output attached. might be that variable names too similar to other variable names. Try running again using design object without other variable columns."))}
  # pick out the columns that are the standard error
  independent.var.column<-1
  dependent.var.n<-(ncol(summary.result.svyby)-1)/2
  stat.columns<-2:(1+dependent.var.n)
  se.columns<-(max(stat.columns)+1):ncol(summary.result.svyby)

  standard_error <- summary.result.svyby[,c(independent.var.column,se.columns)] %>% melt(id.vars=c(independent.var))
  stat <-         summary.result.svyby[,c(independent.var.column,stat.columns)] %>% melt(id.vars=c(independent.var))


  results<-list()
  results$independent.var.value <- stat[,independent.var]
  results$dependent.var.value <- stat[,"variable"]
  results$numbers <-stat[,"value"]
  results$se <- standard_error[,"value"]
  results$min <- results$numbers - results$se
  results$max <- results$numbers + results$se
  return(results)
}





confidence_intervals_num <- function(dependent.var,
                                     independent.var = NULL,
                                     design,
                                     data = data){
  formula_string<-paste0("~as.numeric(", dependent.var, ")")
  summary <- svymean(formula(formula_string), design, na.rm = T)
  confints <- confint(summary, level = 0.95)
  results<-list()
  results$names <- dependent.var
  results$numbers <- summary
  results$min <- confints[,1]
  results$max <- confints[,2]
  return(results)
 }

confidence_intervals_num_groups <- function(dependent.var,
                                     independent.var,
                                     design,
                                     data = data){
  formula_string <- paste0("~as.numeric(", dependent.var,")")
  by <- paste0("~", independent.var, sep = "")
  summary <- svyby(formula(formula_string), formula(by), design, svymean, na.rm = T)
  confints <- confint(summary, level = 0.95)
  results<-list()
  results$names <- dependent.var
  results$numbers <- summary
  results$min <- confints[,1]
  results$max <- confints[,2]
  return(results)
}