README.md

In Benjamin-Vincent-Lab/housekeeping: Basic Debugging and Package Building Tasks

The purpose of this package is to add tools that are needed for basic debugging and package manipulation tasks.

• This package was initially hosted at https://github.com/Benjamin-Vincent-Lab/housekeeping and was moved here on 2/13/2024

Debugging

To debug it's often helpful to pull in the variables of a function into the global environment and work with the lines of a function there without using the debugger. To do this manulally you need to copy the called arguments, delete the commas and then do the same for the default function variables. If you are trying to get 3+ nested functions in, this can be annoying.

These two functions are designed to help with that: function_var_to_env - puts all of the function variables in a specified environment (usually global) to make it as if you are working insdie of the function function_to_function - runs from the beginning of fun_name_1 to the start of fun_name_2 and then runs function_var_to_env on that second function

function_var_to_env

Takes the text used to call a function and calls all of the variables in that function into whatever environment you want, global for me. If any variables weren't used in the call it uses the default function values and puts those in the environment so you can debug the function in the global environment. For example:

fun_a = function(var_a = 10, var_b = "b"){
    print(paste0("var_a: ", var_a))
    print(paste0("var_b: ", var_b))
}  

At this point the variables aren't in the global env

var_a     # Error: object 'var_a' not found
var_b     # Error: object 'var_b' not found

Now assign the function text you want to dig into to a variable and run function_var_to_env fun_a_txt = 'fun_a(var_a=1.5)' function_var_to_env(fun_a_txt)

These variables are now in the global environment.

var_a     # 1.5
var_b     # 'b'

function_to_function

If the example for function_var_to_env was run then we can run all the code up to fun_name_2 and function_to_function on fun_name_2 by using this function:

fun_b = function(var_a=1){
  print("start fun_b ----------")
  print(paste0("var_a: ", var_a))
  print(paste0("changed var_a: "))
  var_a = var_a + 1
  print(paste0("var_a: ", var_a))
  print("running fun_a ----------")
  fun_a(
    var_a = var_a
  )
}

Then run:

function_var_to_env(called_fun = fun_b_txt)
function_to_function(fun_name_1 = "fun_b", fun_name_2 = "fun_a")

Finding script locations

get_script_path

It's often helpful for scripts to know where they are so they can write files relative to their own location. This script does just that. It will work if the script containing get_script_path is sourced from: an Rscript, RStudio console, RStudio source or RStudio. If called from a function it will take the path from the script that called it. Also if it is sourced form a source, it will take the path of the parent source. For use in functions I use it like this:

sample_script_path = get_script_path()
do_something = function(script_path = sample_script_path){
  print(script_path)
}

This way the path will be assigned to the function when the function is sourced.

I used this SO question as a starting point: rscript-determine-path-of-the-executing-script

find_folder_along_path

Goes up path component and looks for the folder_name in each directory.

find_folder_along_path

Goes up path component and looks for the folder_name in each directory.

use_nextflow_wd

Tries to set the wd to the wd found in the Nextflow log.

Package Tools

detach_package

Detaches every instance of a package

remove_package_from_all_libraries

nt

package_is_loaded

Tells if a package is loaded

get_loaded_package_version

nt

matches_loaded_version

Tells if entered version matches that of the loaded version

get_package_version_listed_in_description

nt

assemble_package

Save lots of steps in making new packages Modifies the version listed in the DESCRIPTION file Deletes the tar file from any older packages Runs devtools::document on the package If should_build, builds the package to make tar.gz file * Tries to load the package. If it doesn't work, the DESCRIPTION file goes back to the original version

Assembling this package

In R:

housekeeping::assemble_package(package_name = "housekeeping", my_version = "0.3.6",
  my_dir = "/datastore/alldata/shiny-server/rstudio-common/dbortone/packages/housekeeping")

In bash:

cd /datastore/alldata/shiny-server/rstudio-common/dbortone/packages/housekeeping
my_comment="use_nextflow_wd wasn't being found under housekeeping functions."
git commit -am "$my_comment"; git push origin master
git tag -a 0.3.6 -m "$my_comment"; git push -u origin --tags

In R:

devtools::install_github("Benjamin-Vincent-Lab/housekeeping", ref="0.3.6")

Benjamin-Vincent-Lab/housekeeping documentation built on July 29, 2024, 6:04 a.m.