find_function_1param_value_divideby2: find_function_1param_value_divideby2
In daviddoret/GRCRToolkit: GRC Toolkit

Description Usage Arguments Value Examples

Given f(x) = y where f is a monotonic function, find x for a given y.
Use a naive iterative approach dividing and mulitplying steps by 2 until a solution is found within desired precision.
Non-monotic functions are supported when the search range is limited to a monotonic subset of the function's range.
WARNING: This algorithm is naive, i.e. performance-wise suboptimal.

find_function_1param_value_divideby2(f, y_target_value, x_first_guess = NULL,
  x_first_step = NULL, x_search_limit_min = NULL,
  x_search_limit_max = NULL, y_precision = NULL, max_iteration = NULL,
  verbosity = NULL, ...)

`f`	A function receiving a numeric value as its first parameter followed by ...
`y_target_value`	The output of `f(x)` for which we want to find `x`.
`x_first_guess`	(Conditional) An initial best-effort first guess of what `x` could be.
`x_first_step`	(Conditional) An initial step length to find better better values.
`x_search_limit_min`	(Conditional) A strict limit below which `x` will not be searched.
`x_search_limit_max`	(Conditional) A strict limit above which `x` will not be searched.
`y_precision`	(Conditional) An acceptable degree of precision. Default: .01.
`max_iteration`	(Conditional) The maximum number of search loops performed by the function. This is a safeguard to avoid infinite loops. Default: 2048.
`verbosity`	(Conditional) When `verbosity > 0` detailed information will be output to the console.

Numeric vector. x such that f(x) = y_target_value.

precision <- .0001
y_target <- runif(n = 1, min = -1000000, max = +1000000)
c <- runif(n = 1, min = -1000000, max = +1000000)
m <- runif(n = 1, min = -1000000, max = +1000000)
f <- function(x) { return(c + m * x) }
x_found <- find_function_1param_value_divideby2(
  f = f,
  y_target_value = y_target,
  y_precision = precision,
  verbosity = verbosity)
y_found <- f(x_found)
delta <- abs(y_target - y_found)
label <- paste0(
  "f(x): c + m * x",
  "\nc: ", fn(c,8),
  "\nm: ", fn(m,8),
  "\ny_target: ", fn(y_target,8),
  "\nx_found: ", fn(x_found,8),
  "\ny_found: ", fn(y_found,8),
  "\ndelta: ", fn(delta,8))
message(label)