Significance of Attribute Name of "function evals".

[sundar3492]

Let say, I have well behaved data-set and I try to fitting this data-set with two Gaussian fitting.

Now, I found the below output,

[[Model]]
    (Model(gaussian1) + Model(gaussian2))
[[Fit Statistics]]
    # fitting method   = leastsq
    # function evals   = 29
    # data points      = 100
    # variables        = 6
    chi-square         = 8297.41156
    reduced chi-square = 88.2703358
    Akaike info crit   = 453.852870
    Bayesian info crit = 469.483891
[[Variables]]
    amp1:  122.899857 +/- 1.52937166 (1.24%) (init = 260)
    cen1:  1.00027783 +/- 0.00288716 (0.29%) (init = 1)
    wid1:  0.20129987 +/- 0.00290241 (1.44%) (init = 0.2)
    amp2:  61.0231508 +/- 1.51470406 (2.48%) (init = 140)
    cen2:  1.99916357 +/- 0.00564876 (0.28%) (init = 2)
    wid2:  0.19744994 +/- 0.00567828 (2.88%) (init = 0.19)
[[Correlations]] (unreported correlations are < 0.100)
    C(amp2, wid2) =  0.581
    C(amp1, wid1) =  0.581
    C(wid1, wid2) = -0.103

Now on the [[Fit Statistics]] section of the output what is the Significance of Attribute Name of # function evals = 29

Is it the value of maximum likelihood functions or anything else?

[newville]

The number reported as "# function evals" is the number of times the model function was evaluated in the fit process.

There isn't a single value for "likelihood" reported.