Please help me with an arraylist part.

[bltdavid]

got interrupted by phone call -- first 30 mins coding not done

am interested, so will complete this
will need another 30 mins

I think original author script has some errors

more soon

[mailonline38]

Great!..You are good coder. Yes, the pine script can not plot for high price such as BTCUSD sometime.
....Thanks so much...Take you time, David.

[bltdavid]

coding done, debugging ...

[bltdavid]

Coding done, testing done, but looks useless on a 5min ES chart.

Naturally, I have a feeling there are some calculation errors.

For example, VPCc inside the PriceFunc is poorly described
at the original TradingView website. They imply it is an array,
but I think the 'c' in VPCc could be a typo ... or not.

I had to add some extra code to make sure VPCc and VPRr
were not causing divide by zero -- when the Compare method
shows they're too close to zero, I just set them to 1, which means
they should contribute nothing to the formula (because 1/1 is 1).

Anyways, use the attached code as a basis for continued
experimentation -- I'm not interested in continuing because I
don't know Pine Script -- and like I said I think the original
Pine source is a bit flakey. He has variables lenT and offset,
but never uses them, VPCI is calculated twice, and the biggie
is VPCc looks like an array inside PriceFunc -- arrrgh!

Not knowing Pine Script means my frustration at guessing
at the original code is aggravating, esp when I don't know if
the original code has any bugs or not -- like I said, some things
just look kinda flakey.

But, that could just be my ignorance showing.

Good luck!

[bltdavid]

The script is a lot cleaner, should help you learn quite a bit more
about NinjaScript -- esp with DataSeries.

[mailonline38]

Yeahhhh....Thank you so much, David.
Yes, I did open my view about arraylist and now DataSeries.

I know the logic after converting into NT7, it have a bit different plot as the picture.

[bltdavid]

Looks pretty close!

[mailonline38]

Thanks a lot again. I am studying your code for further understanding. You made my day.

[bltdavid]

I am reading about Series<double> now to avoid using ArrayLists.

Oh .. ok ... well, uh, just be aware ... using Series<double> is NT8.

I bring this up because this is an NT7 forum.

NT8 has made many code-breaking changes. For NT8, they opted to use
a generic type parameter when creating their new Series class, so in NT8
we use Series<double> ... which was formerly the DataSeries class in NT7.

In NT8, Series<int> is same as IntSeries class from NT7.
Series<bool> is same as BoolSeries class.
Series<string> is same as StringSeries class.
etc

The concepts between the two are very similar.
Just realize there are many syntax differences between NT7 and NT8.

An NT7 script cannot run as-is on NT8.

All NT7 scripts must be modified at the source code level to accommodate
the many changes made to the NT8 NinjaScript framework.

-=o=-

Saying "data series" in NT7 usually meant the "DataSeries class" or the
"IDataSeries interface" -- it was relatively specific. The "IDataSeries
interface" was kinda like the gateway to understanding what the term
"data series" really meant. It was a nice bridge to understanding what
the other series named things were, like a bar series, an IntSeries, etc.

In NT8, saying "data series" has started evolving into a more generic term,
and now usually refers to anything using the new Series<T> class -- where
type T is your "data" when speaking of an NT8 "data series".

My point is, in NT7 a "data series" was talked about in both generic and
specific ways, and you had to infer from context what was really being
said. In NT8, to my way of thinking, the term "data series" now leans
heavily generic, and much less specific. This is a good thing.

-=o=-

I should say, however, that NT8 NinjaScript includes a new method called
AddDataSeries, which I think is a very poor name -- it should have been
called AddBarSeries, since that better fits what it actually does.

My point is: regulating the term "data series" to the completely generic term
it really is didn't happen completely in NT8. They embedded the term into a
tangential relationship with creating a new bar series -- a mistake, I think -- esp
when you consider a fantastic name like AddBarSeries was readily available.

Oh well ... I guess, at the time, they forgot to ask me ... LOL



EDIT: In NT8, when selecting the bar series for the chart, the dialog that
allows you to do this is titled "Data Series". It seems NT folks are trying to
redefine "data series" to mean "bar series" -- a direction that seems
completely unnecessary. Data series should be the generic description,
just like the Series class has become generic. A bar series is a specific
instance of a data series, but NT pollutes their own terminology.

[bltdavid]

Anyways, use the attached code ...

I found a bug.

Inside the PriceFunc method, you'll see this loop
at the bottom,

double Price = 0;
for (int i = 0; i < lenV; ++i)
{
    double VPRr = VPR[i];
    if (VPRr.Compare(0, 0.000000000001) == 0)
        VPRr = 1;
    Price += Low[B][COLOR=#e74c3c][0][/COLOR][/B] * 1/VPCc * 1/VPRr;
}

The Low[0] should be Low[i].

My bad.

[mailonline38]

I found a bug.

Inside the PriceFunc method, you'll see this loop
at the bottom,

double Price = 0;
for (int i = 0; i < lenV; ++i)
{
double VPRr = VPR[i];
if (VPRr.Compare(0, 0.000000000001) == 0)
VPRr = 1;
Price += Low[B][COLOR=#e74c3c][0][/COLOR][/B] * 1/VPCc * 1/VPRr;
}

The Low[0] should be Low[i].

My bad.

Yes, it looks better...Thanks so much David.

[mailonline38]

Hello @bltdavid,
I have learn from the clean way of your code with DataSeries and PrintString a lot. Thanks so much.
The tryAVSL script plotted is fine but it can not plot for any security with the price above 10,000 scale.
Do you have any hint or tips for this problem? Please advise.

[bltdavid]

Not really, but I'll take a look ...

What instrument are you testing with?
YM, NQ ... ??

[mailonline38]

Not really, but I'll take a look ...

What instrument are you testing with?
YM, NQ ... ??

Yes, YM, NQ and RTY with Timeframe 60min and below (ES is OK). Thanks for your reply.

[bltdavid]

I have no idea.

I've done my best interpretation, trying to faithfully translate the Pine
script directly to NinjaScript.

I've read through the Pine script again, and even reading through the
user comment section on the tradingview.com website.

The original pine script has issues, some have been acknowledged in
the comments.

The main issue, I think, is no one seems to have (or understand) the
correct formula. The source of the original algorithm is a book, and
complaints in the user comment section imply the book author has
explained this algorithm poorly.

I agree, something is wrong with my translation. But I don't know what
it is, and I don't know where to look to find an answer.

Unfortunately, you're on your own from here.

Good luck!