# Tom Lane

### MathWorks

Statistics and Machine Learning Toolbox developer

Professional Interests: statistics, especially reliability/survival analysis, design of experiments, anova, simultaneous inference, visualization

**Statistics**

RANK**89**

of 244,323

REPUTATION**1,186**

CONTRIBUTIONS

**0** Questions

**428** Answers

ANSWER ACCEPTANCE **
0.00%
**

VOTES RECEIVED**215**

RANK

of 17,075

REPUTATION**N/A**

AVERAGE RATING**0.00**

CONTRIBUTIONS**0** Files

DOWNLOADS **0**

ALL TIME DOWNLOADS**0**

CONTRIBUTIONS**0 Posts**

CONTRIBUTIONS**0** Public Channels

AVERAGE RATING

CONTRIBUTIONS**0** Highlights

AVERAGE NO. OF LIKES

**Content Feed**

How to fit a multi modal distribution using a weighted sum of PDFs?

I'm glad Bjorn provided an answer that works for you. For future reference, there is a function for fitting mixtures of normal d...

8 months ago | 1

Nonlinear fit to multiple data sets with shared parameters

@Tai-Yen Chen, if you concatenate, then the sharing is built in. You don't have to account for different subsets of the data fol...

9 months ago | 0

Warning: Maximum likelihood estimation did not converge. Function evaluation limit exceeded.

Since the random numbers vary, it makes sense that the number of iterations will also vary so the number of warnings may change....

1 year ago | 1

| accepted

Nonlinear fit to multiple data sets with shared parameters

My posted answer from 11-Dec-2012 shows how to deal with multiple datasets following roughly the same functional form with param...

1 year ago | 0

Nonlinear fit to multiple data sets with shared parameters

My posted answer from 11-Dec-2012 shows how to deal with multiple datasets that are roughly similar. That isn't necessary, thoug...

1 year ago | 0

Curve fitting: seversl curves to one

If you want nhh = a * (n1^b) * (n2^c) * (n3^d) consider taking logs log(nhh) = log(a) + b*log(n1) + c*log(n2) + d...

1 year ago | 0

Do I purchase Statistics and Machine Learning Toolbox separately from MATLAB license?

There may be other options. If the GUI author can work with you, you may be able to replace use of tdfread by a newer core MATLA...

3 years ago | 0

Understanding the Stepwiselm PRemove

Daria, thanks for providing the data, allowing me to reproduce your results. It looks like the documentation is confusing or ...

3 years ago | 0

| accepted

How to obtain Std of Coefficients from Curve Fitting

You can get more information when you invoke the fit command: [obj,gof,opt] = fit(...) This gives the fitted obj, goodne...

3 years ago | 2

Nonlinear Regression with Errors in X and Y

If you just had y and one or more x variables as predictors, there is information about an errors-in-variables fit here: <htt...

4 years ago | 0

Kernel Density estimation with chosen bandwidth, then normalize the density function (cdf) so that integral of cdf from min to max equal to 1 ; then take the first and second derivative of the cdf

You seem to want to do a number of things including integrating and specifying a bandwidth. Maybe this will get you started. ...

4 years ago | 0

Unable to Generate Code out of Machine Learning Model

Your screen shot shows generating MATLAB code from the app. That code does training. I think you want to export a compact model ...

4 years ago | 0

How to perform stratified 10 fold cross validation for classification in MATLAB?

If you have the Statistics and Machine Learning Toolbox, consider the |cvpartition| function. It can define stratified samples.

4 years ago | 1

| accepted

Conflicting results with multcompare when using the Kruskal-Wallis test on multiple groups

It's sad but true that there can be an overall difference according to one test, another test might not declare specific differe...

5 years ago | 0

Log likelihood for each distributions.

It's not clear to me what fails to match with what. These match: >> x = -5*log(rand(100,1)); >> pd = fitdist(x,'weibull'...

5 years ago | 1

| accepted

bayesian logistic regression - slicesample - finding Machine learning parameters

It looks like you have the right idea. But I suspect the calculations are underflowing. You're multiplying thousands of probabil...

5 years ago | 1

how to identify a fitglm output as being rank deficient from the resulting object

If your model is f, you could see if f.NumCoefficients > f.NumEstimatedCoefficients

5 years ago | 0

How to calculate the correlation coefficient between an array and a matrix?

If you have the Statistics and Machine Learning Toolbox, it sounds like you want this: >> x = randn(20,3); >> y = x*[1 0...

5 years ago | 0

| accepted

Unable to access Stats toolbox functions

The toolbox has functions anova1, anova2, anovan, and rmanova among others. It does not have functions anova or ranova. Howev...

5 years ago | 1

plotSlice - what are the numbers below the plots?

The prediction shown at the left of the plot is the value given by the model when the predictors are set to the numbers shown be...

5 years ago | 3

| accepted

Does stepwisefit function is able to evaluate using adjusted R-squared instead p-value?

There is a newer function that can do that: load hald stepwiselm(ingredients,heat,'Criterion','adjrsquared')

5 years ago | 1

| accepted

Why is fitlm (or regess) and estimation using mathematical equations giving different results?

The first two columns of coefficients have what appear to be exact zeros in row 13, corresponding to column 12 of X because of t...

5 years ago | 0

What is the difference between the regress function and the fitlm function

Take a look at the 12th and 13th columns of X. It looks to me like the 12th may be constant or may differ by a constant from the...

5 years ago | 0

Decision trees, only binary branches?

If you fit a classification tree to the famous Fisher iris data, you get this: >> load fisheriris >> f = fitctree(meas,s...

5 years ago | 0

| accepted

How can I make DOE design?

Here are two ways to do it. First, specify that you have 5 factors a-e and you want a resolution 5 design. Resolution 5 means yo...

5 years ago | 1

Separate Drawing of Gaussian Mixture Model

You did something like this: x = [randn(4000,1)/2; 5+2*randn(6000,1)]; f = fitgmdist(x,2); histogram(x,'Normalization...

5 years ago | 2

| accepted

clustering, matlab, nominal data

For hierarchical clustering, consider using Hamming distance. Here's an example that isn't realistic but that illustrates what t...

6 years ago | 0

Why confidence interval distributions overlap the distribution?

A couple of things. First, take out your truncate statements and run the code. You'll see that the "LB" cdf is substantially bel...

6 years ago | 0

fitnlm w/ table using not all Variables

You can specify modelfun using variable names: load carsmall t = table(MPG,Weight,Origin) nlm = fitnlm(t,'MPG~b1+b2*W...

6 years ago | 0

| accepted

Dimensions Reduction in Matlab using PCA

The first component explains most of the variation in the columns of DATA, but Y is not involved in that. Of course I don't unde...

6 years ago | 1

| accepted