sequential integer random seed starts at

[MartaVT]

Hello,
I am running STRUCTURE 2.3.4 and I am concerned about the meaning of the new option "sequential integer random seed starts at".
I read in other posts that: you would use the same seed number only if you want to repeat the analysis in question. Typically, it is best to either randomize seed, or if you are doing a large set of batch files that are being run by a script, you can give each batch file a different random seed. The key is that your seeds need to be different values to make sure you are giving the program the ability to search through a broader area of parameter space. (Otherwise, your repeated runs are really just running the exact same run 5 times in a row.).
Therefore I assume I should use as different (random) seeds as possible in my analyses, right?
Should that be the case, which is the use of the "sequential integer random seed" option. Any hint? (I cannot find it in the manual, sorry, maybe I am missing something).
Thanks for your time and help.
Marta

[Vikram Chhatre]

Hi Marta,

By default STRUCTURE randomizes seed.  In the commandline version, this is automatically set with the option RANDOMIZE=1.  Are you encountering this option when setting up parameters?

V

--
You received this message because you are subscribed to the Google Groups "structure-software" group.
To unsubscribe from this group and stop receiving emails from it, send an email to structure-softw...@googlegroups.com.
To post to this group, send email to structure...@googlegroups.com.
Visit this group at http://groups.google.com/group/structure-software.
For more options, visit https://groups.google.com/groups/opt_out.

[MartaVT]

Hello Vikram,

thanks for taking some time to answer this. I am working with a Windows version. I get this option when starting a new job (Structure Scheduler window). Right below the setting the number of K to explore.

Best,

Marta

El lunes, 16 de diciembre de 2013 15:57:26 UTC+1, Vikram Chhatre escribió:

Hi Marta,

By default STRUCTURE randomizes seed.  In the commandline version, this is automatically set with the option RANDOMIZE=1.  Are you encountering this option when setting up parameters?

V

On Mon, Dec 16, 2013 at 4:36 AM, MartaVT <marta.vil...@gmail.com> wrote:

Hello,
I am running STRUCTURE 2.3.4 and I am concerned about the meaning of the new option "sequential integer random seed starts at".
I read in other posts that: you would use the same seed number only if you want to repeat the analysis in question. Typically, it is best to either randomize seed, or if you are doing a large set of batch files that are being run by a script, you can give each batch file a different random seed. The key is that your seeds need to be different values to make sure you are giving the program the ability to search through a broader area of parameter space. (Otherwise, your repeated runs are really just running the exact same run 5 times in a row.).
Therefore I assume I should use as different (random) seeds as possible in my analyses, right?
Should that be the case, which is the use of the "sequential integer random seed" option. Any hint? (I cannot find it in the manual, sorry, maybe I am missing something).
Thanks for your time and help.
Marta

--
You received this message because you are subscribed to the Google Groups "structure-software" group.

To unsubscribe from this group and stop receiving emails from it, send an email to structure-software+unsub...@googlegroups.com.

[Nima Shariati]

Hey everyone

I'm new in Structure and I have problem to understand some option and terms. What's the meaning of the new option "sequential integer random seed starts at" in Structure V. 2.3.4? When should I use this option? Besides what's the meaning of SEED? Please someone advise.

[Pablo G Goikoetxea]

Hi Marta

If you run Structure just once, then you need to set up only one seed (random or not). But if you run it sequentially (for different Ks) you need a seed for each. By ticking that box you provide sequential seeds (starting at the number you choose) for all your runs.

Hope this helps

Pablo

[f.pina...@gmail.com]

Hi Nima,
Please refer to Pablo's excellent answer regarding the "sequential integer random seed starts at" option.
Here is a little background on the "why" of using these SEED values, and how they work.

Francisco