Circular_Genome-mutationPointOffsetRange causes segfault

[VincentRagusa]

In circular genome:

mutationPointOffsetRange = 1.0 #(double) range of PointOffset mutation

the default value of the parameter is 1.0 and is described as a double value, but the value 0.1 causes a segfault.

[cliff-bohm]

Yea, I'm aware there is an issue there. Contracts, you are the first person to use this who is not me. Someone needs to take a look and make a fix.

…

On Mon, Jul 13, 2020 at 12:03 PM Vincent Ragusa ***@***.***> wrote: In circular genome: mutationPointOffsetRange = 1.0 #(double) range of PointOffset mutation the default value of the parameter is 1.0 and is described as a double value, but the value 0.1 causes a segfault. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#346>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABXPY6QSQPW4ET5ODFGWNCTR3MV3VANCNFSM4OYUCONA> .

-- Cliff Bohm - Evolving Systems Technologist BEACON Center for the Study of Evolution in Action Michigan State University

[cliff-bohm]

... and by "someone", I mean I'll take a look...

…

On Mon, Jul 13, 2020 at 1:08 PM Clifford Bohm ***@***.***> wrote: Yea, I'm aware there is an issue there. Contracts, you are the first person to use this who is not me. Someone needs to take a look and make a fix. On Mon, Jul 13, 2020 at 12:03 PM Vincent Ragusa ***@***.***> wrote: > In circular genome: > > mutationPointOffsetRange = 1.0 #(double) range of PointOffset mutation > > the default value of the parameter is 1.0 and is described as a double > value, but the value 0.1 causes a segfault. > > — > You are receiving this because you are subscribed to this thread. > Reply to this email directly, view it on GitHub > <#346>, or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ABXPY6QSQPW4ET5ODFGWNCTR3MV3VANCNFSM4OYUCONA> > . > -- Cliff Bohm - Evolving Systems Technologist BEACON Center for the Study of Evolution in Action Michigan State University

-- Cliff Bohm - Evolving Systems Technologist BEACON Center for the Study of Evolution in Action Michigan State University

[cliff-bohm]

I take it back. I ran a new million mutations without segfault... I'm not seeing any issue here. Are you using an up to date version of the code? If so, I'm going to need a test case that fails (at least settings_organism.cfg that crashes).

I do remember there being an issue at one point. It's possible that the issue was fixed, and I forgot that I fixed it already...

[VincentRagusa]

This file recreates it on both master and development. Sorry if I'm missing something obvious.
settings_organism.txt

[cliff-bohm]

Okay, I think I see what's up. You have sitesType set to int. set it to double and see if you get better results. an offset of < 1 with sites type int is garbage. it would not crash, so I'll need to look into that.

…

On Tue, Jul 14, 2020 at 10:46 AM Vincent Ragusa ***@***.***> wrote: This file recreates it on both master and development. Sorry if I'm missing something obvious. settings_organism.txt <https://github.com/Hintzelab/MABE/files/4919441/settings_organism.txt> — You are receiving this because you were assigned. Reply to this email directly, view it on GitHub <#346 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABXPY6XYGQGC5MU6A7NAY7TR3RVTTANCNFSM4OYUCONA> .

-- Cliff Bohm - Evolving Systems Technologist BEACON Center for the Study of Evolution in Action Michigan State University

[cliff-bohm]

er "should not", not "would not"

…

On Tue, Jul 14, 2020 at 10:49 AM Clifford Bohm ***@***.***> wrote: Okay, I think I see what's up. You have sitesType set to int. set it to double and see if you get better results. an offset of < 1 with sites type int is garbage. it would not crash, so I'll need to look into that. On Tue, Jul 14, 2020 at 10:46 AM Vincent Ragusa ***@***.***> wrote: > This file recreates it on both master and development. Sorry if I'm > missing something obvious. > settings_organism.txt > <https://github.com/Hintzelab/MABE/files/4919441/settings_organism.txt> > > — > You are receiving this because you were assigned. > Reply to this email directly, view it on GitHub > <#346 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ABXPY6XYGQGC5MU6A7NAY7TR3RVTTANCNFSM4OYUCONA> > . > -- Cliff Bohm - Evolving Systems Technologist BEACON Center for the Study of Evolution in Action Michigan State University

-- Cliff Bohm - Evolving Systems Technologist BEACON Center for the Study of Evolution in Action Michigan State University

[VincentRagusa]

Right of course. That makes sense. Thanks for taking a look