scalanlp / junto Goto Github PK

This project forked from parthatalukdar/junto

This toolkit provides an implementation of Modified Adsorption (MAD), a graph-based semi-supervised learning (SSL) algorithm.

Home Page: https://github.com/scalanlp/junto

License: Apache License 2.0

Scala 100.00%

junto's People

Contributors

Stargazers

Watchers

Forkers

dhgarrette younes-abouelnagah math4youbyusgroupillinois phiradet kogecoo

junto's Issues

Enable publishing of artifacts using bintray.

E.g., based on the setup used for lib-text:

https://github.com/peoplepattern/lib-text

Get prepositional phrase attachment test case working.

The existing test case is now broken and should be fixed before any further changes.

Use Scallop instead of Argot for command line parsing.

ScalaNLP is standardizing(ish) around using Scallop across projects, and anyway it is way better than Argot.

Make the simple example work again.

Make it possible to run it as a standalone example.

Restructure the package a bit, e.g. put the graph stuff in junto.graph.

save graph using utf-8 encoding

GraphIo only uses the default encoding, which is bad for data that requires utf-8

Running the simple example gives MRR and Accuracy as 0.0

This is after fixing issue #18

Number of nodes: 4
Number of edges: 5
1: delta:0.5491498679251984 obj:2.1960180733906567
2: delta:0.10661104947846078 obj:2.01739037613283
3: delta:0.04077357380979312 obj:1.9000600041927704
4: delta:0.02015839412263533 obj:1.8626257274679043
5: delta:0.011605686605780988 obj:1.8464512420420467
6: delta:0.0065220398546859525 obj:1.8351359645562653
7: delta:0.0037211986986449342 obj:1.830067299741951
8: delta:0.0021141925168267142 obj:1.8267260410463952
9: delta:0.0012035394144654606 obj:1.825039556510467
TIME: 0.014
Accuracy: 0.0
MRR: 0.0

Cannot build on scala 2.11.x

I noticed that this project cannot build on Scala 2.11.x when I tried it with below command:

$ sbt +package

It produced UNRESOLVED DEPENDENCY error during fetching the library org.rogach.scallop.
It seems that the scallop library for Scala 2.11 is distributed from 0.9.5 or later see mvnrepository.
But Junto depends on 0.9.4, so it fails to compile on Scala 2.11.

The quick & minimal fix is of course modify scallop version from 0.9.4 -> 0.9.5.

eval-label-file should be an optional argument

Right now, omitting it causes a match error.

Exception in thread "main" scala.MatchError: None (of class scala.None$)
	at junto.Junto$.main(Junto.scala:29)
	at junto.Junto.main(Junto.scala)
	at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
	at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
	at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
	at java.lang.reflect.Method.invoke(Method.java:498)
	at com.intellij.rt.execution.application.AppMain.main(AppMain.java:147)

Use sbt-native-packager.

Need a way of creating an executable w/ all the jars and such sorted out automatically.

Minimal example as test case.

Make the minimal example work as a test case, similar to PrepAttachSpec.

Bonus: make the evaluation code a bit cleaner so that one isn't pulling out the label associations out of LabelSpecs, etc.

specifying eval-label-file argument causes index out of bounds exception

command line args to reproduce

-e examples/minimal/edges.minimal.tsv  -i 10 -o examples/minimal/predictions.tsv -s examples/minimal/seed_labels.minimal.tsv -t -v --eval-label-file examples/minimal/reference.output_predictions.minimal.csv

Exception in thread "main" java.lang.ArrayIndexOutOfBoundsException: 1
	at junto.graph.package$$anonfun$getLabels$2.apply(package.scala:48)
	at junto.graph.package$$anonfun$getLabels$2.apply(package.scala:44)
	at scala.collection.Iterator$$anon$11.next(Iterator.scala:409)
	at scala.collection.Iterator$$anon$13.hasNext(Iterator.scala:462)
	at scala.collection.Iterator$$anon$11.hasNext(Iterator.scala:408)
	at scala.collection.Iterator$class.foreach(Iterator.scala:893)
	at scala.collection.AbstractIterator.foreach(Iterator.scala:1336)
	at scala.collection.TraversableOnce$class.toMap(TraversableOnce.scala:316)
	at scala.collection.AbstractIterator.toMap(Iterator.scala:1336)
	at junto.Junto$.main(Junto.scala:36)
	at junto.Junto.main(Junto.scala)