{"id":4046,"date":"2021-04-01T08:16:52","date_gmt":"2021-04-01T12:16:52","guid":{"rendered":"https:\/\/blogs.mathworks.com\/loren\/?p=4046"},"modified":"2021-04-01T10:14:32","modified_gmt":"2021-04-01T14:14:32","slug":"having-an-argument-in-release-r2021a","status":"publish","type":"post","link":"https:\/\/blogs.mathworks.com\/loren\/2021\/04\/01\/having-an-argument-in-release-r2021a\/","title":{"rendered":"Having an Argument in Release R2021a"},"content":{"rendered":"
Have you ever looked at code where you are calling a function with many arguments, many of which are strings, and find it hard to see what's going on? I know I have. And perhaps you too. In release <\/span>R2021a<\/span><\/a>, there is a new feature in the language that can help me out, in addition to the relatively new <\/span>arguments<\/span> block which I wrote about <\/span>here<\/span><\/a>. For making your own argument, you might want to watch <\/span>this classic video<\/span><\/a>.<\/span><\/div>

Make a plot to modify<\/span><\/h2>
Let's start by making a plot and annotating it.<\/span><\/div>
t = 0:0.1:10;<\/span><\/span><\/div><\/div>
y = sin(2*pi*60*t) + 0.2*randn(size(t));<\/span><\/span><\/div><\/div>
plot(t,y)<\/span><\/span><\/div><\/div>
xlabel(<\/span>\"Time (in seconds)\"<\/span>)<\/span><\/span><\/div><\/div>
ylabel(<\/span>\"Signal (in Voltage)\"<\/span>)<\/span><\/span><\/div><\/div>
title(<\/span>\"Canonical signal\"<\/span>)<\/span><\/span><\/div><\/div>
grid(<\/span>\"on\"<\/span>)<\/span><\/span><\/div>
<\/div><\/div><\/div><\/div>
<\/div><\/div><\/div>

Change the plot via name-value pairs (2020 and before)<\/span><\/h2>
Now let's say I want to recreate the same plot but with more control over the appearance. I could use name-value pairs, which we've been using in MATLAB for years.<\/span><\/div>
plot(t,y,<\/span>\"Marker\"<\/span>,<\/span>\"o\"<\/span>,<\/span>\"MarkerEdgeColor\"<\/span>,<\/span>\"k\"<\/span>,<\/span>\"MarkerFaceColor\"<\/span>,<\/span>\"y\"<\/span>,<\/span>\"LineStyle\"<\/span>,<\/span>\"-.\"<\/span>,<\/span>\"LineWidth\"<\/span>,1.0)<\/span><\/span><\/div><\/div>
xlabel(<\/span>\"Time (in seconds)\"<\/span>)<\/span><\/span><\/div><\/div>
ylabel(<\/span>\"Signal (in Voltage)\"<\/span>)<\/span><\/span><\/div><\/div>
title(<\/span>\"Discrete signal\"<\/span>)<\/span><\/span><\/div><\/div>
grid(<\/span>\"on\"<\/span>)<\/span><\/span><\/div>
<\/div><\/div><\/div><\/div><\/div>
As you read your way through that first line to <\/span>plot<\/span> right above, it's easy to get lost, especially if the values happen to be longer strings.<\/span><\/div>

Change the plot via name=value (starting in R2021a)<\/span><\/h2>
I could use the new feature in R2021a to use <\/span>name=value syntax<\/span><\/a> (as well as <\/span>here<\/span><\/a>) to accomplish the same result.<\/span><\/div>
plot(t,y,Marker=<\/span>\"o\"<\/span>,MarkerEdgeColor=<\/span>\"k\"<\/span>,MarkerFaceColor=<\/span>\"m\"<\/span>,LineStyle=<\/span>\":\"<\/span>,LineWidth=2.0)<\/span><\/span><\/div><\/div>
xlabel(<\/span>\"Time (in seconds)\"<\/span>)<\/span><\/span><\/div><\/div>
ylabel(<\/span>\"Signal (in Voltage)\"<\/span>)<\/span><\/span><\/div><\/div>
title(<\/span>\"Discrete signal\"<\/span>)<\/span><\/span><\/div><\/div>
grid(<\/span>\"on\"<\/span>)<\/span><\/span><\/div>
<\/div><\/div><\/div><\/div><\/div>

Smooth the data<\/span><\/h2>
Now let's try smoothing that signal and display the results.<\/span><\/div>
smoothedData = smoothdata(y,<\/span>\"movmean\"<\/span>,SmoothingFactor=0.5,SamplePoints=t);<\/span><\/span><\/div><\/div>
plot(t,y,<\/span>\"Color\"<\/span>,[77 190 238]\/255,DisplayName=<\/span>\"Input data\"<\/span>)<\/span><\/span><\/div><\/div>
hold <\/span>on<\/span><\/span><\/div><\/div>
plot(t,smoothedData,Color=[0 114 189]\/255,LineWidth=1.5,<\/span>...<\/span><\/span><\/div><\/div>
DisplayName=<\/span>\"Smoothed data\"<\/span>)<\/span><\/span><\/div><\/div>
hold <\/span>off<\/span><\/span><\/div><\/div>
legend<\/span><\/span><\/div><\/div>
xlabel(<\/span>\"Time (in seconds)\"<\/span>)<\/span><\/span><\/div><\/div>
ylabel(<\/span>\"Signal (in Voltage)\"<\/span>)<\/span><\/span><\/div><\/div>
title(<\/span>\"Smoothed signal\"<\/span>)<\/span><\/span><\/div><\/div>
grid(<\/span>\"on\"<\/span>)<\/span><\/span><\/div>
<\/div><\/div><\/div><\/div><\/div>

Discussion<\/span><\/h2>
This syntax has some restrictions, listed in the release notes. E.g., the <\/span>name=value<\/span> syntax is only valid for trailing in inputs. In other words, once you opt in to the syntax for a given function call, all the remaining inputs must also be in <\/span>name=value<\/span> form. Since the syntax is new, there are also some affordances missing for the moment, like tab-completion. From my perspective, I think it makes my code even more self-documenting than before, since the burden of matching with string inputs is no longer an issue. What are your thoughts on <\/span>name=value<\/span> syntax for optional, trailing input arguments? Let us know <\/span>here<\/span><\/a>.<\/span><\/div>
Copyright 2021 The MathWorks, Inc.<\/span><\/div>\r\n<\/div>