In Rkabacoff/qacStats: Statistical functions for QAC courses

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

csv_file_table <- function(path, bin_var, nbins=10, bp=NULL, adherentvar=NULL, export_file_name){
  require(dplyr)
  require(readxl)

  #Reading in dataset from Excel (.xls or .xlsx), turning dataset to dataframe
  dataset <- read_excel(path, na=c("", "NA"))
  dataset <- as.data.frame(dataset)

  #Binning variable
  bin_var <- as.numeric(as.character(dataset[[bin_var]]))
  if(!is.null(bp)){
    bp <- bp
  } else{
    bp <- pretty(bin_var, n=nbins)
  }
  y <- cut(bin_var, breaks=bp, labels=NULL)

  #Summary table w/ dplyr
  y <- enquo(y)
  results = dataset %>%
    group_by(!!y) %>%
    summarise(n=n()) %>%
    mutate(prop=n/sum(n)) %>%
    mutate(cum_freq=cumsum(n)) %>%
    mutate(cum_prop=cumsum(prop))

  #Adding adherent var to summary table w/ dplyr
  if(!is.null(adherentvar)){
    adherentvar <- enquo(adherentvar)
    adherent_results = dataset %>%
      group_by(!!adherentvar) %>%
      summarise(n=n()) %>%
      mutate(prop=n/sum(n))
    results = cbind(results, adherent_results)
  }

  #Final .CSV file
  write.table(results, file=export_file_name, sep=",", row.names=FALSE)
}

line_plot <- function(path, primaryaxis, secaxis){
  #Reading in .csv file, converting to a dataframe
  results_csv <- read.csv(path)
  results_csv <- as.data.frame(results_csv)

  #Renaming first column to intervals, omitting missing values, & turning 'intervals' column from factor to numeric
  names(results_csv)[1] = "intervals"
  results_csv = results_csv[results_csv$intervals != "NA",]
  results_csv <- na.omit(results_csv)

  #Plot
  primaryaxis <- deparse(substitute(primaryaxis))
  secaxis <- deparse(substitute(secaxis))
  par(mar = c(5, 5, 3, 5))
  plot(results_csv[[primaryaxis]], type ="l", ylab = "Cumulative Proportion",
       main = "Plot of Cumulative Proportion and Proportion", xlab = "Intervals",
       col = "blue")
  axis(side=4)
  par(new = TRUE)
  plot(results_csv[[secaxis]], type = "l", xaxt = "n", yaxt = "n",
       ylab = "", xlab = "", col = "red", lty = 2)
  mtext("Proportion", side = 4, line = 3)
  legend("topleft", c("Cumulative Proportion", "Proportion"),
         col = c("blue", "red"), lty = c(1, 2), cex = 0.4)
}

Function Summaries

The first function of this package, csv_file_table, generates a CSV file with the summary table of the binned variables. For the csv_file_table function, type path = file.choose() and click on the Excel file (.xls or .xlsx) that contains the dataset. bin_var refers to the variable you want to bin - make sure argument is a string. You can switch the count of nbins you desire - the default is set to 10. You can identify the intervals you want in bp or you can allow pretty() within the function to create 'pretty' intervals. If you have an adherent variable that you want to be part of your summary table you can indicate it as an argument. export_file_name takes in a string input for what you want the file name of the summary table to be. The function produces a .csv file with the summary table results in your working directory.

The second function of this package, line_plot, generates a line plot from the summary table produced in the first function. For line_plot function, type path = file.choose() and click on the .csv file with the summary table results created through the csv_file_table function. primaryaxis is the variable you want plotted on the left y axis. secaxis is the variable you wanted plotted on the right y axis.

Examples

In the first example, a CSV file was created with the summary statistics for the variable days_to_follow_up1.

path = "/Users/belen/Desktop/senior/Advanced R/sample_data.xls"

csv_file_table(path=path, bin_var="days_to_follow_up1", export_file_name="results.csv")

Below is the CSV file created with the summary statistics for the variable days_to_follow_up1.

In the second example, the cumulative proportion and proportions calculated within the CSV file's summary table in the first function is plotted, in the primary and secondary axis respectively, with the X axis as the binned intervals.

path = "/Users/belen/results.csv"

line_plot(path=path, primaryaxis=cum_prop, secaxis=prop)

Rkabacoff/qacStats documentation built on Jan. 17, 2024, 9:25 p.m.