Zorro 1.36

Posted By: jcl

Zorro 1.36 - 09/29/15 10:13

A new beta version was uploaded:

http://server.conitec.net/down/ZorroBeta.zip

V1.36 is a slightly newer version with implemented features for updating variables in external text files, and exporting balance and equity curves.
Posted By: Sphin

Re: Zorro 1.36 - 09/29/15 21:54

Code:
with implemented features for updating variables in external text files
Almost a bit creepy, but I'm just looking for such a possibility. Is there a doc available yet?

BTW: it would be nice to find the new TO_CSV print option in the list of print parameters, I surely can guess about the form but how to tell the command to which CSV file it should write?
Posted By: jcl

Re: Zorro 1.36 - 09/30/15 07:02

It's under "printf". The file name is the name of the script plus extension ".csv".

You can find all new commands also under "What's New" in the manual.
Posted By: Sphin

Re: Zorro 1.36 - 09/30/15 08:51

Quote:
It's under "printf". The file name is the name of the script plus extension ".csv".

Ah! It's in the help file of 1.34 only, not in the online manual. I thought the online manual is the most actual version. laugh

Quote:
You can find all new commands also under "What's New" in the manual.

There's nothing about a function to update variables in external text files so far but I guess it's coming soon, no problem.

Thanks!
Posted By: GPEngine

Re: Zorro 1.36 - 10/02/15 14:22

I get a script crash in Bootstrap.c from http://www.financial-hacker.com/boosting-systems-by-trade-filtering/

The crash doesn't happen if I comment out the line
Code:
randomize(BOOTSTRAP|DETREND,Sample,Curve[j].Values,Curve[j].Length);
Posted By: GPEngine

Re: Zorro 1.36 - 10/02/15 15:31

It is because you have defined

var Sample[3000];

My curves are 3040 long. I guess I get more days than you for the same StartDate = 2010;
Posted By: GPEngine

Re: Zorro 1.36 - 10/02/15 15:32

I mean, this is a classic source of bugs in C code. It is sharply painful when trying to run code someone else wrote....
Posted By: jcl

Re: Zorro 1.36 - 10/05/15 08:42

Yep. Static arrays are ugly hacks and should not really be used in commercial quality code. On the other hand, they are temptingly easy to define.
Posted By: Sphin

Re: Zorro 1.36 - 10/11/15 00:51

Are there new rules for Date intervals since 1.36.4?

e.g.

Code:
BarPeriod = 5;
StartDate = 20140106;
EndDate = 20140131;
vars Price = series(price());


results in a 041 Inconsistent series calls, while

Code:
BarPeriod = 5;
StartDate = 20140101;
EndDate = 20140114;
vars Price = series(price());


works. In 1.36.1 this problem did not appear.

Update:

Even the Z strategies break with "041 Inconsistent series calls" in 1.36.4.
Posted By: jcl

Re: Zorro 1.36 - 10/12/15 06:51

Yes, this was a bug in the previous upload of 1.36. Please download the beta again - it should be fixed.
Posted By: Sphin

Re: Zorro 1.36 - 10/12/15 17:42

This issue has been fixed, thanks!

I don't know if you have seen my entry about the new issue with different results when using a TMF (http://www.opserver.de/ubb7/ubbthreads.php?ubb=showflat&Number=455104#Post455104). It also happens in 1.36.5, at least Stop and TakeProfit are effected when calling a TMF, they seem to be modified by a hidden mechanism although they are left untouched by script. So IMO you actually can not rely on results when involving TMFs.
Posted By: jcl

Re: Zorro 1.36 - 10/12/15 17:54

Ah, no, I had not seen it. Of course, this script will produce a random result. You're not returning anything from your TMF, so it just closes or continues the trade randomly, depending on the printf return value.

A TMF that does nothing must return 0. Omitting the return value won't work.
Posted By: Sphin

Re: Zorro 1.36 - 10/16/15 00:37

The print command returns any value other than zero? Okay, I did not consider this. Setting explicitly 'return 0' at the end of the TMF produces same results as without the TMF. Thanks!
© 2024 lite-C Forums