How to calculate mean square error, i'm using mean(x) function but its giving NaN value. i have also attached my code below.

clear all; clc;


load('UDDSprofile.mat')
Time = UDDSprofile(2:end,2)' - UDDSprofile(2,2);
U = UDDSprofile(2:end,3)';
Vt_exp =  UDDSprofile(2:end,4)';


ode_fun=@(t,y) BATTERY(t,y,Time,U);
[T,Y] = ode45(ode_fun,[0 3500],[0.7 0 0]);
x1=Y(:,1); x2=Y(:,2); x3=Y(:,3);

u = interp1(Time,U,T);

load('HPPC_parameters.mat');
r0 = interp1(SOC,R0,x1);
V_out =-r0.*u+ boc(x1)-x2-x3;
V_exp=interp1(Time,Vt_exp,T);

%RMSE
E = (V_exp - V_out);
SE = E.^2;
MSE = mean(SE);
RMSE = sqrt(mean((V_exp - V_out).^2));

[Walter Roberson]

Unfortunately we do not have your mat files so we cannot test.

u = interp1(Time,U,T);

Instead of doing that, you can pass in U as the tspan (provided it is in sorted order) instead of passing in [0 3500] . When you pass in a vector with more than 2 entries as the tspan then MATLAB will report back for exactly those locations only, making it unnecessary to interpolate afterwards.

[the cyclist]

It can be a bit cavalier to immediately leap to the

mean(y,'omitnan')

solution, without understanding the reason why the NaNs are in the data. It seems more likely to me that the existence of NaNs in SE (and therefore in E, and therefore in either V_exp or V_out, etc.) is the problem.