In Stat-Cook/Data.Quality.Reports:

knitr::opts_chunk$set(echo=F, warning=F, message = F)

{
  library(tidyverse)
  library(lubridate)
  library(caret)
  library(pROC)
  library(glue)
  library(NURS.Data.Quality)
}

style_table <- function(data){
  DT::datatable(
    data,
    options = list(scrollX = TRUE)
  )
}

``` {r, echo=F}

comp <- function(values, limit){ freq <- table(values) any(freq >= limit * sum(freq)) }

freq_limit <- function(data, limit){ results <- blank_result(data) for (name in names(data)) { values <- data[, name] results[name] <- comp(values, limit) } results }

freq_frame <- function(data){ tibble( Variable = colnames(data), 100% Modal Class = freq_limit(data, 1), >90% Modal Class = freq_limit(data, 0.9), >80% Modal Class = freq_limit(data, 0.8), >50% Modal Class = freq_limit(data, 0.5) ) } hr <- params$human.readable

## `r hr`

### Overview

``` {r}

data <- readRDS(params$data.location)
data %>% head() %>% style_table

Singular Summary

Rate at which most common (modal) value dominates variable:

freq_frame(data) %>% style_table()

Missing Summary

Rate at which values are missing:

tibble(
  Variable = colnames(data),
  `>10% Missing` = average_missing_heuristic(data, 0.1),
  `>30% Missing` = average_missing_heuristic(data, 0.3),
  `>50% Missing` = average_missing_heuristic(data, 0.5),
  `>90% Missing` = average_missing_heuristic(data, 0.9)
)%>%  style_table()

Stat-Cook/Data.Quality.Reports documentation built on May 4, 2022, 2:21 p.m.