NOTE: This is the fifth in my series of “10 things I wish I had known when I started using MATLAB”.

This is one of those tips that’s not really a tip at all. It’s more like an admission of guilt and ignorance. I hope you learn from my mistakes!

Name your variables something meaningful

It’s so obvious that it hardly bears mentioning, but … you should always try to name your variables with meaningful names. This will help you in so many ways, especially as your script develops and your workspace gets more and more crowded (side note: read my post about avoiding workspace “bloat”).

For example, maybe we have a vector of names, and instead of calling it something generic, like data or vec or, my personal favorite, temp (because I’m lazy), let’s call it something that will remind us and/or inform someone else reading our script what the variable contains, like I don’t know … names:

names = ["Gary", "Donna", "Richard", "Antonia"];

That wasn’t so hard, was it?!?!

Beware the punctuation

Sometimes, though, it can be easy to get carried away when naming variables and you start using punctuation or spaces. Well, to be technically accurate, you can’t use punctuation or spaces in variable names because MATLAB won’t let you. For example, maybe we want to be really descriptive and call our vector of names first names; well, that’s not allowed. MATLAB will only allow us to name variables according to these rules:

  • the variable name must start with a letter
  • the variable name can be followed by letters, digits or underscores
  • the letters are case sensitive, so “a” is different than “A”

So, we could name our vector first_names (notice the underscore in place of the space).

We cannot name our vector first_names! or first_names* or anything else containing punctuation (other than the underscore character).

Super beware the punctuation in tables

I know, “super beware” is not a phrase, but maybe it will catch on. Anyhow, what about tables? Tables are great because we can assign a name to each column, and the variable types represented in each column can be different. So, we have a lot of flexibility in how we refer to our data (just remember that in a table each column must have the same number of rows).

Here’s an example of a table with measurements of atmospheric particulate matter (PM) concentration made with different instruments (that we saw in a previous post about using flags):

waitbar

We can access data from this table by referring to row and/or column names. For example, if I want the set of measurements from the reference instrument (“PM_ref”):

waitbar

Very handy because I don’t have to remember that “PM_ref” is the fourth column. In fact, it could be moved around within the table and I could still refer to it as PA_PM_data.PM_ref because it would retain that name.

However, I ran into trouble when naming my columns because I was trying to be too descriptive and precise. Instead of calling the column “PM_ref”, I called it “PM 2.5 ref”. Why did I do this you ask? Because technically the quantity being measured is the particulate matter concentration of aerosol particles less than 2.5 $\mu m$ in diameter - hence the “2.5” in the name. And the real problem is that MATLAB let me do it!

waitbar

Now I can’t refer to the “PM 2.5 ref” column as easily with PA_PM_data.PM 2.5 ref because my column name has spaces in it and a period! Doh!

I can still access that column by name, but it’s a lot clunkier:

meas = PA_PM_data{:,'PM 2.5 ref'};

So, just don’t use spaces or punctuation in your table names even though it is allowed. Just don’t do it!