Maximal number of iterations allowed for finding a smoothing spline has been reached

[HealthyPear]

Sometimes during gta.optimize() I get this warning,

The maximal number of iterations maxit (set to 20 by the program)
allowed for finding a smoothing spline with fp=s has been reached: s
too small.
There is an approximation returned but the corresponding weighted sum
of squared residuals does not satisfy the condition abs(fp-s)/s < tol.
  spline = UnivariateSpline(x, y, k=2,

which seems to be triggered within utils.get_parameter_limits().

I noticed that even in the tutorials this warning seems to pop up, but nothing is explained regarding it.

Is this something the user should worry about? If yes, is there a recommended way to approach the issue e.g. using the available keyword arguments of gta.optimize()?

Slightly unrelated, but still, unexpected behavior: while debugging this I noticed that the optimizer section of my config is not respected by this method when I override one key, as I have

gta.config["optimizer"] = {'optimizer': 'MINUIT',
 'tol': 0.001,
 'max_iter': 100,
 'init_lambda': 0.0001,
 'retries': 3,
 'min_fit_quality': 2,
 'verbosity': 0}

but the output of gta.optimize(optimizer={"verbosity":3}) shows

Minuit fit quality: 3

I would expect that only verbosity should have been overridden.

I get the same behavior even passing the full dict,

optimizer_dict = {**gta.config["optimizer"], "verbosity": 1}
print(optimizer_dict)
gta.optimize(optimizer=optimizer_dict)

gives

2025-11-25 10:37:39 INFO    GTAnalysis.optimize(): Starting
{'optimizer': 'MINUIT', 'tol': 0.001, 'max_iter': 100, 'init_lambda': 0.0001, 'retries': 3, 'min_fit_quality': 2, 'verbosity': 1}

[...]

Minuit fit quality: 3   estimated distance: 2.25552e-07
Minuit parameter uncertainties:
  1  0.0239165
  2  0.0166402
  3  0.0985877
  4  0.00766259
  5  0.00911272
[...]/lib/python3.11/site-packages/fermipy/utils.py:785: UserWarning: 
The maximal number of iterations maxit (set to 20 by the program)
allowed for finding a smoothing spline with fp=s has been reached: s
too small.
There is an approximation returned but the corresponding weighted sum
of squared residuals does not satisfy the condition abs(fp-s)/s < tol.
  spline = UnivariateSpline(x, y, k=2,