tabular files

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "README-"
)

Guesses the formatting of tabular/flat files by testing different options for formatting the delimiter, decimal mark, grouping mark and column header. Chooses the combinations that creates a data frame that make most sense, based on the assumptions:

That data is rectangular, delimiters that fails to read some of the columns are skipped
That formatting that gives more total number of columns is more likely to be right
That formatting that gives more non-character columns is more likely to be right
That if having columns header doesn't do any difference, then it more likely that headers are not present

For delimiters following are possible: tab (\t), comma (,), semicolon (;) and whitespace/fixed width. As decimal mark: comma (,) and dot (.). As big number grouping mark: comma (,), dot (.) and space ( ) are tested. Column headers existence is also tested for, altogether 22 possible formatting combinations are tested.

This Shiny app https://rokka.shinyapps.io/smartep demonstrates how this package (together with https://github.com/ijlyttle/shinypod) can be used to create an tabular/flat file reader that you can throw almost any file at - and it will in most cases guess the right formatting for you.

Quick start

Install

Dev version from GitHub.

devtools::install_github("lukas-rokka/readrGuess")

library(readr)
library(readrGuess)

Example

# Print all formatting combinations that are currently tested
delim_cases()

# Create a date frame and format it into a semicolon delimited string
test_str <- readr::format_delim(
  data.frame(a=runif(1000, -100, 100), b="a", c=1, d=TRUE), 
  ";")

# Read the string and guess the right formatting
read_guess(test_str)