Same strategy, different ninjascript, different results

[bltdavid]

I am learning via Microsoft virtual school, they have great tutorials.

Excellent.

[NGorodzhiy]

Great thanks! Will do!

[sledge]

That's because the two pieces of code are not the same.

Remove the semi-colons from the end of the conditionals of all your if statements.
.

Nice eye. I've caught those before -- but was on my Samsung Galaxy Note 3 trying to figure this out at the time.

But the one still has rsi[0], and the other script is RSI(2, 0)[0]. (looking at the original green/red code posting).

I haven't tested - but I'm not sure that would return the same exact results.

[NGorodzhiy]

Same results, already tested. My friend declared before that the RSI = 2.

[bltdavid]

I do. CrossAbove and CrossBelow is his version of ">" and "<".

As stated already, CrossAbove is like ">" but it's actually more than that. (Make no mistake, it is definitely not the same as ">").

Let's examine how they are different.

If you say,

if (Close[0] > SMA(20)[0])
{
}

it's obviously "true" for all bars where Close[0] is greater than the value returned by SMA(20)[0]. Makes sense, right?

But CrossAbove() does an additional check, and thus is not the same as the plain vanilla ">" check by itself. This additional check is where CrossAbove's lookback argument comes into play.

Consider that this code,

if (CrossAbove(Close, SMA(20), 1))
{
}

is equivalent to this expanded version,

if (Close[0] > SMA(20)[0] && Close[1] <= SMA(20)[1])
{
}

because the point of CrossAbove is to return "true" only on the bar where the cross over actually happens.

See the difference?

Using ">" returns true on the crossover (of course) but it also returns true on all bars *after* the crossover where the condition is still true.

Using CrossAbove is more like a finely tuned version of ">" because it only returns true on the *actual* bar (just one bar, not 2, not 3, just 1) where the cross over *actually* occurs, all the remaining bars after the cross over return "false".

CrossAbove returns false for bars after the cross over because of that second check where it looks back 1 bar to see if the values *on that bar* were below.

Using ">" returns "true" repeatedly whereas CrossAbove returns "true" just once.

Make sense?

[bltdavid]

Same results, already tested. My friend declared before that the RSI = 2.

On your next post, attach the actual files.

I want to see 2 attachments that are the actual two files that you're complaining about.

[NGorodzhiy]

Thank you, now I understand. What exactly am I "complaining" about?

[bltdavid]

Thank you, now I understand. What exactly am I "complaining" about?

Isn't this thread devoted to the lack of consistent results between 2 strategies that you say are functionally the same?

Therefore, isn't the "complaint" the fact the results are *not* the same?

So, please attach the 2 files that are producing these inconsistent results.

Your "complaint" or "issue" (consider it legal-speak) can be resolved faster if you'd just attach the actual 2 files causing the "issue".

Otherwise, from my perspective, I have insufficient data to help you solve your complaint, er, I mean, problem.

Look, everyone here helping you (except for Ninja folks) is volunteering their time.

Help us to help you (it would go sooooo much faster) if you would just attach the 2 files causing the problem.

Otherwise, if you can't do that (sometimes pro's ask for help here but they can't upload any code because of internal rules at their company), well, I can't read your mind and I can't download/read/investigate code I can't see, so then I'm done being able to help.

[NGorodzhiy]

The problem is already solved haha. You helped me solve it with your first post, and I appreciate it a lot. Thank you.

[bltdavid]

Same results, already tested. My friend declared before that the RSI = 2.

Sorry, my mistake.

I mistook this to say "same incorrect results" ... thus I felt like we needed to see your actual files to investigate further ...

My apologies. Glad to hear it's working!