How to limit a calculation for an array for when the cells in an array are greater than 0

I'm looking apply an equation and calculate an output - The equation is:

IE = 1367*(1+0.033 .* cos(0.0172024.*DN(i))) .* sin(SOLALT(i));

where DN and SOLALT are both 105120x1 arrays.

I want to use the calculation for IE when SOLALT>0, and set IE = 0 when SOLALT <= 0

how do I achieve this?

I have tried to create a for loop but have been unsuccessful.

for i = 1:length(x)
if SOLALT(i) > 0   
IE = 1367*(1+0.033 .* cos(0.0172024.*DN(i))) .* sin(SOLALT(i));
end
end
plot(IE)

x a variable with the value 105120

This is the code I attempted

[Cris LaPierre]

What is EI? Is that supposed to be IE?

[Scott Muir]

Sorry,

IE and Ei were the same. This was a mistake.

Answer 1

[Walter Roberson]

Logical indexing.

IE = zeros(size(SOLALT));
mask = SOLALT > 0;
IE(mask) = 1367*(1+0.033 .* cos(0.0172024.*DN(mask))) .* sin(SOLALT(mask));

Answer 2

[Cris LaPierre]

You don't need a for loop. Try using a logical array (see Ch 12 of MATLAB Onramp).

IE = 1367*(1+0.033 .* cos(0.0172024.*DN)) .* sin(SOLALT);
IE(SoLALT<=0)=0;

[Walter Roberson]

Cris's method can be very effective. The difference between Cris's approach and what I posted is that Cris's approach calculates everywhere and then replaces the unwanted parts, whereas mine does not calculate for the unwanted parts. This has several consequences:

• if most of the outputs are unwanted, the expense of processing everything can potentially be high
• but it costs to work conditionally, so especially if most of the outputs are wanted, it can be less expensive to replace only the unwanted ones on output
• the processing could have side effects (like incrementing counters or writing to files), and sometimes it is important that those side effects not take place for the unwanted locations
• in some cases, processing at the unwanted locations can create error() conditions, so sometimes it is important to only proceed for the wanted locations