Eigenvector University Europe Returns to Rome, October 12-15, 2026

Cleve Moler

Jun 14, 2026

Cleve Moler passed away May 20, 2026 and, quite rightly, much has been said about his contributions to the fields of mathematics and scientific computing. As creator of MATLAB his work had a profound impact on millions of scientists and engineers. The MATLAB universe includes tools for users in many different disciplines, from automobile and process control to finance. I’ve been using MATLAB for 40 years now and of course all of Eigenvector’s software products are built with it, including our for-MATLAB products like PLS_Toolbox and stand-alone products like Solo. It greatly accelerated the field of chemometrics, chemical data science and machine learning. 

But for me, the significance of his work can be shown in a simple example. Last week a question came up about degrees of freedom in multivariate calibration. Just to make sure I was thinking correctly on why you have to account for mean centering as consuming a degree a freedom I created this example. 

MATLAB commands illustrating that mean centering data consumes a degree of freedom.

Cleve gave us the tools to do linear algebra calculations easily in a language that felt natural. It’s hard to overstate what this did for education in this field. The ability to simply try things to confirm understanding is huge. And writing MATLAB code was so much easier because you didn’t have to pre-allocate memory or multiply matrices in loops and it gave sensible error messages making it easy to debug. My first day working with MATLAB was my last day writing code in BASIC or FORTRAN. In my earliest days in chemometrics I’d hear a talk at a conference about a new idea or method and then go back to my room and program it up. MATLAB made data science so much fun! Exciting times. 

I am fortunate to have had a few interactions with Cleve. One of the earliest, in 1995, involved a discussion on how multi-way operations, i.e. tensor factorizations and multiplications might be built into MATLAB. The last, in 2024, had to do with singular matrix pencils in the generalized eigenvalue problem Ax = lambda Bx. This is a problem sometimes in the Generalized Rank Annihilation Method (GRAM). In a note about this to our Eigenvector staff Cleve added:

Let me tell you: when Cleve Moler asks if you are interested, there is only one answer!

Many years ago we gave Cleve one of our shirts with the Eigenvector logo embroidered on it. He said, “I get lots of swag, but I’ll actually wear this.” It makes me happy to think that maybe he did. Given the work he did on LINPACK, EISPACK and eigenfunctions of vibrating membranes, I think he found it amusing that somebody would start a company called “Eigenvector.” 

Cleve Moler’s work had an enormous influence on me; without it Eigenvector Research might not exist. That impact flows to my fellow Eigenvectorians and the thousands of users of our software. Rest in peace Cleve. Thank-you for everything, or as we say here at Eigenvector, EVRIthing. 

BMW