What makes the residual table have NaN values?

I am fitting a model (15 observables) to experimental data on Simbiology using Particle Swarm Optimization and a pooled fit, to optimize the values of 10 parameters in the model. However, no matter what I try with the initial values, the residual table generated has a lot of NaN values, and then I can`t plot the bootstrap confidence intervals (although I am able to obtain the Gaussian CI). So, what makes that happen? When I look at the fitted model responses, they seem to be close to the experimental data points, so I am not sure why the residuals are returned as NaN. Any ideas?

Thanks!

[Torsten]

When I look at the fitted model responses, they seem to be close to the experimental data points, so I am not sure why the residuals are returned as NaN.

If the fitted model responses are all near to the experimental data points and the experimental data points have no NaN values, the residuals cannot have NaN values because they are defined as experimental data points - fitted data points.

[Rebeca Hannah Oliveira]

Thanks!

[the cyclist]

This is not easy to help you debug without seeing the data and code. Can you upload them? You can use the paper clip icon in the INSERT section of the toolbar. The ideal is to share the minimal example that illustrates the problem.