RE: ILD script

[Ângelo P. Pinto]

Hi Ross,

Thank you for your messages. No, I don’t have any difficult in to manage TNT use of computer memory. However I have observed some weird behaviors with the amount of memory ram used and number of maxtrees saved, but it is other subject.

Probably I will need understand deeply the macro language (and even the ILD test, two tasks which I was avoiding to do now) to solve my problems. Anyway I separated the dataset in pairs of blocks and the analysis worked.

Best wishes,

Ângelo

---

Date: Fri, 31 May 2013 04:11:24 -0700
From: ross....@gmail.com
To: tnt-tree-analysis-u...@googlegroups.com
CC: jgomes_...@yahoo.com.br
Subject: Re: ILD script

A related error message you might also get if dealing with a large amount of data is

"Not enough RAM available to continue..."

if so, you'll need to increase the available RAM with the mxram command
http://tnt.insectmuseum.org/index.php/Commands/mxram

Ross

On Monday, 27 May 2013 19:28:31 UTC+1, Ângelo P. Pinto wrote:

Dear Ross,

Thank you very much for your help. Considering that some people could be having the same problem I posted a feedback here.

I have solved the problem depicted before, copying and pasting the full "ild.run" script in the same txt file containing the script for read the matrix, as well as the searching parameters. I also included the blocks specifications in the end of matrix, great suggestion. I did few tests with small hypotheticals matrices (such as that above) containing at maximum three blocks and shazam! The files ran and all quotations written in the ild script appeared. However when I tested with a real matrix containing one morphological block (only discrete; num) and four more blocks with molecular data (dna), summing up 4089 characters the script stops to read in the line 29 depicting a problem with the same command line “var: numblocks chararray[(nchar 1)] truelength thislength critical charnum p maxiter;”... I will check more carefully but in advance I would like to tell you that the problems are not with blocks partitions, since TNT has identified the blocks correctly.

Furthermore even when the script runs correctly there are some errors reports:

1. Warning: Numbers of loops/variables must be changed under "macro-" Request to change numbers of loops/variables will be ignored 

2. Sometimes the window status appears after the test already be performed and stays freeze, although the program still working correctly.

All the best,
Ângelo

p.s. I did a test reading the file "ild.run" separately just after the matrix be read and also works without any problem…

On Saturday, May 25, 2013 10:21:47 AM UTC-3, Ross Mounce wrote:

Is it really necessary specifying blocks in files already arranged in blocks of different data types (e.g. num; DNA etc.)? 

No.  It might be easier to specify your blocks at the end of your datafile like this:

blocks 0 61;

(the first block is characters 1 to 62, the second block is implicit characters 63 to last)

Then just:

log output.log;

proc yourdata.tnt; 

xmult:level10; 

sect:slack 40;

run ild.run;

Ross

--
Has recibido este mensaje porque estás suscrito a un tema del grupo "TNT-Tree Analysis using New Technology" de Grupos de Google.
Para anular la suscripción a este tema, visita https://groups.google.com/d/topic/tnt-tree-analysis-using-new-technology/FB66wgdjlwQ/unsubscribe?hl=es. Para anular la suscripción a este grupo y todos sus temas, envía un correo electrónico a tnt-tree-analysis-using-n...@googlegroups.com.
Para publicar una entrada en este grupo, envía un correo electrónico a tnt-tree-analysis-u...@googlegroups.com.
Visita este grupo en http://groups.google.com/group/tnt-tree-analysis-using-new-technology?hl=es.
Para obtener más opciones, visita https://groups.google.com/groups/opt_out.
 
 

[Mike Noren]

For people who come here in the future, here's how to get the ILD script to work:

1) Start TNT, and under Settings turn OFF "Enable Macros".
2) Under Settings, select memory. Set General RAM to some reasonable value (e.g. 512), and under Macros increase KBytes and Variables (for my 77 taxon, 7000 bp dataset 2048 KBytes and 10000 variables worked).
3) Load your dataset.
4) Set the New Technology Search options you want.
5) On the command line, define your data blocks by indicating the starting position of each data segment, separated by space, with the BLOCKS command (e.g. BLOCKS 0 651 3761)
6) On the command line, type sect:slack 128 and press return.
7) Now you can start the ILD script by typing ild on the command line. You will probably need more than the default 100 replicates, try e.g. ild 500

Hopefully this may help someone, especially as it also gets the other scripts to work. It took me almost three hours to figure out.




On Saturday, April 20, 2013 12:03:46 AM UTC+2, Ângelo P. Pinto wrote:

Hi all, 

How execute an ILD test with the Siddall's script using a mixed data set with several blocks of molecular {DNA} and morphology {NUM} arranged in the same ".txt" file? Apparently is necessary editing some parameters in script file before running the analysis, what are these parameters? I checked the history of topics in the group and there is at least one topic with same question dated from October of 2012, but without answer. Maybe someone can help.

Matrixes something like this:

xread
15 4 
&[num] 
Taxon_1 11[03]11 
Taxon_2 1[23][03]10 
Taxon_3 1[23][03]10 
Taxon_4 1[12][03]10 
&[dna] 
Taxon_1 CCATCTACAA 
Taxon_2 CCATCTACAA 
Taxon_3 CCATCTACAA 
; 
proc/; 

Thank you in advance.
Best wishes, 
Ângelo