Options

It is possible to pass an optional set of parameters to the fitting functions with the options keyword, using an Options struct. For example:

using EasyFit, Random
Random.seed!(1)

x = sort(rand(10))
y = @. 0.3 * exp(5x)

fitexp(x, y, options=Options(maxtrials=1000))

------------ Single Exponential fit -----------

Equation: y = a exp(-x/b) + c

With: a = 0.3000000000000478
      b = -0.2000000000000075
      c = -1.326315096604623e-13

Correlation coefficient, R² = 1.0
Average square residue = 1.0136631520496675e-26

Predicted Y: ypred = [ 0.33840859531570533, 0.3836158209060227, ...]
residues = [ -8.021361352916756e-14, -7.499556531342932e-14, ...]

-----------------------------------------------

Available options:

Keyword	Type	Default value	Meaning
`fine`	`Int`	100	Number of points of fit to smooth plot.
`p0_range`	`Vector{Float64}`	`[-100(maximum(Y)-minimum(Y)), 100(maximum(Y)-minimum(Y))]`	Range of generation of initial random parameters.
`nbest`	`Int`	5	Number of repetitions of best solution in global search.
`besttol`	`Float64`	1e-4	Similarity of the sum of residues of two solutions such that they are considered the same.
`maxtrials`	`Int`	100	Maximum number of trials in global search.
`debug`	`Bool`	false	Prints errors of failed fits.

The nbest, besttol, and maxtrials options are also used by the multistart search performed by fitexpdecay (see Normalized exponential decay).