# Multiple Linear Regression using fitlm function

9 views (last 30 days)
ANURAG DEEPAK on 18 Jan 2020
Answered: Image Analyst on 18 Jan 2020
Hello Sir, why i am not getting the intercept for the other variables?
>> lm=fitlm(X,TAG)
Warning: Regression design matrix is rank deficient to within machine precision.
> In classreg.regr.CompactTermsRegression/checkDesignRank (line 35)
In LinearModel.fit (line 1237)
In fitlm (line 121)
lm =
Linear regression model:
y ~ 1 + x1 + x2 + x3 + x4 + x5 + x6 + x7
Estimated Coefficients:
Estimate SE tStat pValue
__________ __ _____ ______
(Intercept) 0 0 NaN NaN
x1 0 0 NaN NaN
x2 0.00037516 0 Inf NaN
x3 0.00021467 0 Inf NaN
x4 -0.16078 0 -Inf NaN
x5 0.68268 0 Inf NaN
x6 -0.0013354 0 -Inf NaN
x7 0 0 NaN NaN
Number of observations: 5, Error degrees of freedom: 0
F-statistic vs. constant model: NaN, p-value = NaN

Star Strider on 18 Jan 2020
In a linear regression of any sort, there is only one intercept.

ANURAG DEEPAK on 18 Jan 2020
Then sir how to deal with such situation, 7 independent variable and 1 dependent variable?
Star Strider on 18 Jan 2020
If you want to ssee what the intercepts of the individual variables are, you need to regress them individually. However, the entire idea of a multiple linear regression is to regress the various predictor variables together, to get a unified idea of how they all interact. Note that they do not have to be individual predictors, and can be interaction terms (predictors multiplised together) that distorts the idea of an intercept.
If you want to see what variables best predict the dependent variable, use the stepwiselm function.