errors concept

concepts/errors/.meta/config.json

@@ -0,0 +1,5 @@
+{
+  "authors": ["colinleach"],
+  "contributors": [],
+  "blurb": "R has a rich but relatively unfamiliar condition handling system."
+}

concepts/errors/about.md

@@ -0,0 +1,85 @@
+# About
+
+As explored in the `nothingness` concept, R has various ways to represent data which is absent (`NULL`), unknown (`NA`) or mathematically invalid (`NaN`, `Inf`).
+In general, it will try to use these to flag problems in-place, and continue without raising an exception.
+
+## `stop()`
+
+Some problems are too fundamental to let the program continue. In this case, use `stop(msg)` to immediately halt the program with an `Error` and print `msg`.
+
+```R
+> f <- function() {
++     stop("I have a problem")
++ }
+> f()
+Error in f() : I have a problem
+```
+
+## `stopifnot()`
+
+Similar to an `assert` in other languages, this is a convenient wrapper for `stop()`. 
+The parameters are an arbitrary number of boolean conditions, and the program will only continue if they are all `TRUE`.
+
+```R
+> stopifnot(1 < 2) # continues OK
+> stopifnot(1 < 2, 1 == 2)
+Error: 1 == 2 is not TRUE
+```
+
+## `try()`
+
+R has a sophisticated variety of ways to handle errors.
+These are based on Common Lisp and differ significantly from many popular languages.
+
+The simplest is to wrap potentially problematic code in `try()`
+
+```R
+g <- function(val) {
+  log10(val)
+}
+
+> g(3)
+[1] 0.4771213
+
+> g("3")
+Error in log10(val) : non-numeric argument to mathematical function
+
+# with try() and a default value
+h <- function(val) {
+  result <- NA
+  try(result <- log10(val))
+  result
+}
+
+> h(3) # works as before
+[1] 0.4771213
+
+> h("3") # returns a default value plus the error message
+Error in log10(val) : non-numeric argument to mathematical function
+[1] NA
+```
+
+## `warning()` and `message()`
+
+These are non-fatal conditions, to warn of a problem that the program could at least partly recover from, or inform the user of progress in a long-running program.
+
+One difference between them is that warnings are cached and (by default) output when the function returned, but messages are output immediately.
+There are many ways to change the default behavior, beyond our scope here.
+
+```R
+w <- function() {
+  warning("something strange happened")
+  -1
+}
+
+> w()
+[1] -1
+Warning message:
+In w() : something strange happened
+
+# if warnings are not wanted
+> suppressWarnings(w())
+[1] -1
+```
+
+The `suppressWarnings()` function can occasionally be useful within Exercism, to avoid confusing the test runner with unwanted output.

concepts/errors/introduction.md

@@ -0,0 +1,80 @@
+# Introduction
+
+As explored in the `nothingness` concept, R has various ways to represent data which is absent (`NULL`), unknown (`NA`) or mathematically invalid (`NaN`, `Inf`).
+In general, it will try to use these to flag problems in-place, and continue without raising an exception.
+
+## `stop()`
+
+Some problems are too fundamental to let the program continue. In this case, use `stop(msg)` to immediately halt the program with an `Error` and print `msg`.
+
+```R
+> f <- function() {
++     stop("I have a problem")
++ }
+> f()
+Error in f() : I have a problem
+```
+
+## `stopifnot()`
+
+Similar to an `assert` in other languages, this is a convenient wrapper for `stop()`. 
+The parameters are an arbitrary number of boolean conditions, and the program will only continue if they are all `TRUE`.
+
+```R
+> stopifnot(1 < 2) # continues OK
+> stopifnot(1 < 2, 1 == 2)
+Error: 1 == 2 is not TRUE
+```
+
+## `try()`
+
+R has a sophisticated variety of ways to handle errors.
+The simplest is to wrap potentially problematic code in `try()`
+
+```R
+g <- function(val) {
+  log10(val)
+}
+
+> g(3)
+[1] 0.4771213
+
+> g("3")
+Error in log10(val) : non-numeric argument to mathematical function
+
+# with try() and a default value
+h <- function(val) {
+  result <- NA
+  try(result <- log10(val))
+  result
+}
+
+> h(3) # works as before
+[1] 0.4771213
+
+> h("3") # returns a default value plus the error message
+Error in log10(val) : non-numeric argument to mathematical function
+[1] NA
+```
+
+## `warning()` and `message()`
+
+These are non-fatal conditions, to warn of a problem that the program could at least partly recover from, or inform the user of progress in a long-running program.
+
+```R
+w <- function() {
+  warning("something strange happened")
+  -1
+}
+
+> w()
+[1] -1
+Warning message:
+In w() : something strange happened
+
+# if warnings are not wanted
+> suppressWarnings(w())
+[1] -1
+```
+
+The `suppressWarnings()` function can occasionally be useful within Exercism, to avoid confusing the test runner.

concepts/errors/links.json

@@ -0,0 +1,6 @@
+[
+  {
+    "url": "https://adv-r.hadley.nz/conditions.html",
+    "description": "Conditions chapter in Advanced R, 2nd Ed"
+  }
+]

config.json

@@ -586,6 +586,11 @@
       "uuid": "2751b6f2-7d71-4397-b063-9bf927a57756",
       "slug": "booleans",
       "name": "Booleans"
+    },
+    {
+      "uuid": "f9bc1f9b-aa22-48c3-bc1d-ad2ea75a4d36",
+      "slug": "errors",
+      "name": "Errors"
     }
   ],
   "key_features": [