Hello Aaron,
The warning message you are seeing indicates that the function `nlinfit` is encountering some issues while fitting the sigmoidal curve to your data. The warnings suggest that the matrix used in the fitting process is rank deficient, meaning that it does not have full rank.
To address this issue, you can try normalizing your data or applying a transformation to make it more suitable for fitting.
Here's an example of how you can normalize your response data before fitting the curve:
% Normalize the response data
normalized_resp = (resp_conc_hed - min(resp_conc_hed)) / (max(resp_conc_hed) - min(resp_conc_hed));
% Fit the sigmoidal curve using nlinfit
fit_params = nlinfit(conc_nm_hed, normalized_resp, @sigmoidal_function, initial_params);
% Plot the fitted curve
x = linspace(min(conc_nm_hed), max(conc_nm_hed), 100);
y_fit = sigmoidal_function(fit_params, x);
plot(conc_nm_hed, normalized_resp, 'o')
hold on
plot(x, y_fit)
xlabel('5HT Dose (nm)','FontSize',20)
ylabel('Normalized Response to Stimuli (\DeltaF / F)','FontSize',20)
legend('Data', 'Fitted Curve')
In this example, `sigmoidal_function` is a custom function that defines the sigmoidal curve model. You can replace it with your own function that represents the model you want to fit.
For more information you can refer following articles for more information:
