Kstar is not defined for some conditions

[quancs]

Hello, I am currently translating the code to pytorch for using GPUs to speed up and for training neural networks (see Lightning-AI/torchmetrics#1792).
In the following code, I found that Kstar is not defined when BW == cutoffs(6) or BW == cutoffs(7) or BW == cutoffs(8) or BW <= cutoffs(5). Is it because that these conditions won't get satisfied anyway?

SRMRToolbox/SRMR.m

Lines 154 to 162 in f8be861

	if (BW > cutoffs(5)) && (BW < cutoffs(6))
	Kstar=5;
	elseif (BW > cutoffs(6)) && (BW < cutoffs(7))
	Kstar=6;
	elseif (BW > cutoffs(7)) && (BW < cutoffs(8))
	Kstar=7;
	elseif (BW > cutoffs(8))
	Kstar=8;
	end

[tiagofalk]

Hi,

Indeed, it should be:

if (BW >= cutoffs(5)) && (BW < cutoffs(6))
Kstar=5;
elseif (BW >= cutoffs(6)) && (BW < cutoffs(7))
Kstar=6;
elseif (BW >= cutoffs(7)) && (BW < cutoffs(8))
Kstar=7;
elseif (BW >= cutoffs(8))
Kstar=8;
end

kstar is only for conditions where BW>=cutoffs(5) so BW <= cutoffs(5) does not get satisfied.

[quancs]

kstar is only for conditions where BW>=cutoffs(5) so BW <= cutoffs(5) does not get satisfied.

So, the situation BW <= cutoffs(5) will never become true, right?