How to make slope 1in fitlm?

Hello there,

Is it possible to force slope to 1 in in the fitlm function.

Many thanks!

Mohammad

[the cyclist]

I think you need to post some more information about what you are trying to do. Are you asking if you can fit a model of the form

y = a + x

and estimate the value of the intercept a, using fitlm?

Answer 1

[the cyclist]

I don't believe you can do it with fitlm, but you can with fitnlm:

% Set seed for reproducibility
rng default

% Make up some data that obeys your model, with a little noise
x = (1:10)';
y = 7 + x + 0.05*randn(size(x));

% Tabulate the data. (fitnlm does not require this, but it is convenient)
tbl = table(x,y);

% Define the fit equation
f = @(F,x) F(1) + x;

% Define starting guess for parameter
b0 = 1;

% Fit the data
mdl = fitnlm(tbl,f,b0)

mdl = 
Nonlinear regression model:
    y ~ F1 + x

Estimated Coefficients:
          Estimate       SE       tStat       pValue  
          ________    ________    ______    __________

    F1     7.0312     0.027984    251.26    1.2752e-18


Number of observations: 10, Error degrees of freedom: 9
Root Mean Squared Error: 0.0885
R-Squared: 0.999,  Adjusted R-Squared 0.999
F-statistic vs. zero model: 2.11e+05, p-value = 5.55e-21

Answer 2

[behnam ashtarayeh]

Thanks for your answer. Yes, I exatly mean what you said. Is it possible to fit a model of form (y = a + x) using fitlm and estimate intercept (a)?

Answer 3

[behnam ashtarayeh]

Thanks a lot!

It solved my problem.