Confused about RSI

[ydcool]

go-talib/talib.go

Lines 2819 to 2840 in cd53a92

	today := 0
	prevValue := inReal[today]
	prevGain := 0.0
	prevLoss := 0.0
	today++
	for i := inTimePeriod; i > 0; i-- {
	tempValue1 = inReal[today]
	today++
	tempValue2 = tempValue1 - prevValue
	prevValue = tempValue1
	if tempValue2 < 0 {
	prevLoss -= tempValue2
	} else {
	prevGain += tempValue2
	}
	}
	prevLoss /= float64(inTimePeriod)
	prevGain /= float64(inTimePeriod)
	if today > 0 {

Hi Mark, just start reading this source codes, and got a little confused about the today variable in func Rsi, it seems that at L2840 the today will be 3 at least because there's an auto-increament operation at L2823 and min value of inTimePeriod is 2, so the else part at L2850 would never be reached. Is that right?

[ydcool]

I also read the RSI in c at https://sourceforge.net/p/ta-lib/code/HEAD/tree/trunk/ta-lib/c/src/ta_func/ta_RSI.c#l358
I'm not quite familar with c , just wonder that could the startIdx be simplified as 0?

[markcheno]

Hi Dominic,
This code is a straight up port from the Java version of TA-Lib. So any peculiarities like that exist in the original version as well. Since I test the output of this version against the original TA-Lib, I'm inclined to leave it alone for now.

[ydcool]

I see. I have read the java version and finally figured it out. The reason that your code got the correct result is the last for loop coverd the caculations, and those dead codes above have no side effects.