Results
Next
Finally, a 2L / Vana Thread (2015-08-10 11:26:45) :
I don't think I've ever looked at this before; this is ridiculous and strange and wonderful.
I don't think I've ever looked at this before; this is ridiculous and strange and wonderful.
owned by Morrígan, last edited 2015-08-10 11:26:45.
RELEASE: Toolbox SCA 0.6.0 :
[ul][li]Allows insertion / epenthesis of segments ([tt]0 > e / #_sC[/tt])[/li]
[li]Logging to file + console[/li]
[li]Improved logging, including line number and file name for the script in which the error was found.[/li]
[li]Now attempts to log all compilation errors encountered in all loaded scripts before quitting[/li]
[li]Added some clarifications to the manual where users indicated some sections were unclear.[/li][/ul]
https://github.com/samanthamccabe/toolb...
[ul][li]Allows insertion / epenthesis of segments ([tt]0 > e / #_sC[/tt])[/li]
[li]Logging to file + console[/li]
[li]Improved logging, including line number and file name for the script in which the error was found.[/li]
[li]Now attempts to log all compilation errors encountered in all loaded scripts before quitting[/li]
[li]Added some clarifications to the manual where users indicated some sections were unclear.[/li][/ul]
https://github.com/samanthamccabe/toolb...
owned by Morrígan, last edited 2015-08-10 08:34:32.
Sound Change Appliers (2015-07-27 11:41:10) :
That's really weird. [tt]toolbox-sca-0.5.0.jar[/tt] is in that folder right? Everything should be right there. it's very strange that it couldn't find it. You could try also running it manually with [code]java -jar toolbox-sca-0.5.0.jar pq_rules.rule[/code] which is basically what the batch file does. Sorry you are having these problems, I've tested this quite extensively, so I really hope I didn't miss something.
That's really weird. [tt]toolbox-sca-0.5.0.jar[/tt] is in that folder right? Everything should be right there. it's very strange that it couldn't find it. You could try also running it manually with [code]java -jar toolbox-sca-0.5.0.jar pq_rules.rule[/code] which is basically what the batch file does. Sorry you are having these problems, I've tested this quite extensively, so I really hope I didn't miss something.
owned by Morrígan, last edited 2015-07-27 11:44:05.
Sound Change Appliers (2015-07-27 10:24:27) :
Hahah I was JUST thinking about that, because I was curious about JS Burke's Proto-Central-Mountain changes (he gave me an old copy of his rules and lexicon designed for mark's SCA) and I was curious how much more concise (and fast) it would be in my format. But I realized a bunch of his rules probably require this type of treatment. I think I know how to do that.
Ok, to run it: you should be able to just navigate to the folder it's in and run it using [code]toolbox.bat myrules.rule[/co...
Hahah I was JUST thinking about that, because I was curious about JS Burke's Proto-Central-Mountain changes (he gave me an old copy of his rules and lexicon designed for mark's SCA) and I was curious how much more concise (and fast) it would be in my format. But I realized a bunch of his rules probably require this type of treatment. I think I know how to do that.
Ok, to run it: you should be able to just navigate to the folder it's in and run it using [code]toolbox.bat myrules.rule[/co...
owned by Morrígan, last edited 2015-07-27 10:29:06.
v0.5.0 Release :
Guess what I finally did this morning
https://github.com/samanthamccabe/toolbox-sca/releases/tag/v0.5.0
Quick rundown:
I fixed a lot of minor bugs. Create shell scripts / batch file to make running it easier. Changed how the program is invoked slightly. Wrote a 12 page manual, provided example files. Rewrote the regular expression engine, changed how normalization works (also, it's turned off by default).
If anyone wants it, I made a syntax hilighter in Sub...
Guess what I finally did this morning
https://github.com/samanthamccabe/toolbox-sca/releases/tag/v0.5.0
Quick rundown:
I fixed a lot of minor bugs. Create shell scripts / batch file to make running it easier. Changed how the program is invoked slightly. Wrote a 12 page manual, provided example files. Rewrote the regular expression engine, changed how normalization works (also, it's turned off by default).
If anyone wants it, I made a syntax hilighter in Sub...
owned by Morrígan, last edited 2015-07-25 13:01:31.
happiness thread (2015-07-02 10:41:47) :
That's awesome! Congratulations~
That's awesome! Congratulations~
owned by Morrígan, last edited 2015-07-02 10:41:47.
happiness thread (2015-06-19 17:25:56) :
Yeah! That's exactly what I'm doing actually, I just need to work on feature weights (or perhaps, probability of feature-transfer, something like that).
My SCA handles metathesis quite nicely, I'll have you know! I need to finish the manual for that too, I've been slogging through it for a month.
Yeah! That's exactly what I'm doing actually, I just need to work on feature weights (or perhaps, probability of feature-transfer, something like that).
My SCA handles metathesis quite nicely, I'll have you know! I need to finish the manual for that too, I've been slogging through it for a month.
owned by Morrígan, last edited 2015-06-19 17:25:56.
happiness thread (2015-06-17 11:54:51) :
Definitively NOT using Levenstein. Each segment in a sequence is represented by a multivalue feature vector. Right now, this is just a multidimensional vector distance weighted by the strength of each feature (they use different scales), but I'd prefer it represented a probability (or -log thereof) that a pair was related) which is obviously more complicated to model.
I don't have metathesis explicitly implemented yet, but the algorithm already has a way of comparing short (length...
Definitively NOT using Levenstein. Each segment in a sequence is represented by a multivalue feature vector. Right now, this is just a multidimensional vector distance weighted by the strength of each feature (they use different scales), but I'd prefer it represented a probability (or -log thereof) that a pair was related) which is obviously more complicated to model.
I don't have metathesis explicitly implemented yet, but the algorithm already has a way of comparing short (length...
owned by Morrígan, last edited 2015-06-17 11:55:39.
happiness thread (2015-06-13 21:23:04) :
i'm afraid.
I'm just using sequence alignment as a first pass to get likely correspondences from cognate pairs (or more, but alignment in n dimensions might not yield any better results than doing all pairs and tree-building). So it seems pretty uncomplicated.
i'm afraid.
I'm just using sequence alignment as a first pass to get likely correspondences from cognate pairs (or more, but alignment in n dimensions might not yield any better results than doing all pairs and tree-building). So it seems pretty uncomplicated.
owned by Morrígan, last edited 2015-06-13 21:28:43.
happiness thread (2015-06-11 15:44:12) :
yeah, I don't get that. we DID have a quantitative methods course, but frankly it was terrible, the professor sucked (he failed to get tenure, ultimately) and the book sucked. Keith Johnson, I think, the orange one that does everything in R and explains nothing.
yeah, I don't get that. we DID have a quantitative methods course, but frankly it was terrible, the professor sucked (he failed to get tenure, ultimately) and the book sucked. Keith Johnson, I think, the orange one that does everything in R and explains nothing.
owned by Morrígan, last edited 2015-06-11 15:44:12.
happiness thread (2015-06-11 10:24:39) :
They depict the way that model parameters impact the performance of the algorithm vs some human-aligned data. Basically all of that involves sequence alignment stuff, and weight coefficients for my feature model.
I was thinking I could make a thread about this over in Terra Firma
They depict the way that model parameters impact the performance of the algorithm vs some human-aligned data. Basically all of that involves sequence alignment stuff, and weight coefficients for my feature model.
I was thinking I could make a thread about this over in Terra Firma
owned by Morrígan, last edited 2015-06-11 10:24:40.
happiness thread (2015-06-08 19:56:33) :
Oh interesting.
sequence alignment isn't the interesting problem here though, for the most part that's probably going to be inferring proto-forms and rules from correspondences. Which I'm able to get fairly handily. Without training on my Chechen-Ingush-Batsbi data, the system was able to pick out correspondence which I know to be correct. What will be interesting is seeing if it can use a (probably statistical) model to infer reasonable ancestor forms, and identify conditioning environ...
Oh interesting.
sequence alignment isn't the interesting problem here though, for the most part that's probably going to be inferring proto-forms and rules from correspondences. Which I'm able to get fairly handily. Without training on my Chechen-Ingush-Batsbi data, the system was able to pick out correspondence which I know to be correct. What will be interesting is seeing if it can use a (probably statistical) model to infer reasonable ancestor forms, and identify conditioning environ...
owned by Morrígan, last edited 2015-06-08 19:56:33.
happiness thread (2015-06-08 17:11:19) :
There are concerns about that. But mostly I'm interesting in using simulated data to examine how things like inherited vocabulary size and overlap between children effect information recoverability.
Though I have suspicions that when it come to actual reconstruction, unsupervised machine learning might be viable.
There are concerns about that. But mostly I'm interesting in using simulated data to examine how things like inherited vocabulary size and overlap between children effect information recoverability.
Though I have suspicions that when it come to actual reconstruction, unsupervised machine learning might be viable.
owned by Morrígan, last edited 2015-06-08 17:11:20.
happiness thread (2015-06-05 09:34:02) :
So, the latest good news is that I implemented a few new gap penalty functions and so far it looks like the Indo-Iranian data is behaving better with a non-negative gap (using a convex gap function). Still, building more sample data will be a big help.
So, the latest good news is that I implemented a few new gap penalty functions and so far it looks like the Indo-Iranian data is behaving better with a non-negative gap (using a convex gap function). Still, building more sample data will be a big help.
owned by Morrígan, last edited 2015-06-05 09:34:02.
happiness thread (2015-06-02 14:24:17) :
Everything basically. I need to prepare some more data...
http://dialawizard.tumblr.com/post/120528229284/dialawizard-related-to-the-other-data-this-is
also this, where the Indo-Iranian proves to be tricky with the current configuration
http://dialawizard.tumblr.com/post/120495109859/ive-made-some-important-progress-in-laying-the
Everything basically. I need to prepare some more data...
http://dialawizard.tumblr.com/post/120528229284/dialawizard-related-to-the-other-data-this-is
also this, where the Indo-Iranian proves to be tricky with the current configuration
http://dialawizard.tumblr.com/post/120495109859/ive-made-some-important-progress-in-laying-the
owned by Morrígan, last edited 2015-06-02 14:24:17.
happiness thread (2015-06-01 10:36:31) :
I worked SO much this weekend, it was great. I made a test-standard data set for Ingush-Chechen-Batsbi, did a lot of refactoring and modularization of my alignment code, as well as setting up the Jenetics library so I can use genetic algorithms to tune my model parameters.
The results so far are fantastic but i have a [i]lot[/i] of work still ahead. One rather interesting result so far is using a gap penalty (at least, a constant one) is actually bad:
[image:http://haedus.org/f...
I worked SO much this weekend, it was great. I made a test-standard data set for Ingush-Chechen-Batsbi, did a lot of refactoring and modularization of my alignment code, as well as setting up the Jenetics library so I can use genetic algorithms to tune my model parameters.
The results so far are fantastic but i have a [i]lot[/i] of work still ahead. One rather interesting result so far is using a gap penalty (at least, a constant one) is actually bad:
[image:http://haedus.org/f...
owned by Morrígan, last edited 2015-06-01 10:36:31.
happiness thread (2015-05-29 09:53:55) :
A bunch of my friends just got their PhDs. ;_;
My research is coming along though. Bit by bit. And i'm working on the SCA manual and examples :S writing a manual is a lot of work.
A bunch of my friends just got their PhDs. ;_;
My research is coming along though. Bit by bit. And i'm working on the SCA manual and examples :S writing a manual is a lot of work.
owned by Morrígan, last edited 2015-05-29 09:53:55.
happiness thread (2015-05-22 19:37:29) :
[quote kodé, Deacon, this fucking hole we call LA]ABD!!! ABD!!![/quote]
That's fantastic, congratulations. A bunch of my friends just defended, and a few more are trying to be ABD by the end of the summer. Almost makes me miss school.
[quote kodé, Deacon, this fucking hole we call LA]ABD!!! ABD!!![/quote]
That's fantastic, congratulations. A bunch of my friends just defended, and a few more are trying to be ABD by the end of the summer. Almost makes me miss school.
owned by Morrígan, last edited 2015-05-22 19:37:29.
happiness thread (2015-05-08 10:47:49) :
Ah, in this case CD as in a long-term interest bearing account. It's my money to begin with. Apparently if you don't do anything with the bank for 5 years, they will think you might be dead and are required by law to report the money as abandoned.
It's boring though, I'm just gonna throw this at my debt. I have to find [i]something[/i] nice I want to buy...
Ah, in this case CD as in a long-term interest bearing account. It's my money to begin with. Apparently if you don't do anything with the bank for 5 years, they will think you might be dead and are required by law to report the money as abandoned.
It's boring though, I'm just gonna throw this at my debt. I have to find [i]something[/i] nice I want to buy...
owned by Morrígan, last edited 2015-05-08 10:47:49.
happiness thread (2015-05-08 09:56:46) :
I've been so busy. Mixed bag I guess. It's finally really spring here. I just can't seem to help being kind of eternally low-energy and a bit melancholic.
I'm seeing my endocrinologist in a week, for the first time in 6 months, and I can release the next version of the SCA as soon as I update the documentation, and I've made some progress on my real research, which is occasionally exciting.
Best of all, I might get $7000 from a CD i forgot I had and was turned over to the gover...
I've been so busy. Mixed bag I guess. It's finally really spring here. I just can't seem to help being kind of eternally low-energy and a bit melancholic.
I'm seeing my endocrinologist in a week, for the first time in 6 months, and I can release the next version of the SCA as soon as I update the documentation, and I've made some progress on my real research, which is occasionally exciting.
Best of all, I might get $7000 from a CD i forgot I had and was turned over to the gover...
owned by Morrígan, last edited 2015-05-08 09:56:46.
Next
303 results.