forked from ricefield/cs161sp12-linkpred
-
Notifications
You must be signed in to change notification settings - Fork 0
ivmreg/cs161sp12-linkpred
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
Setup: - install redis DB, ruby, python (we used python 2.7.2) - for perceptron: install numpy for python - for libSVM: go to libsvm/python and run "make" [requires g++] - copy libsvm/ to "/usr/libsvm/" Running the pipeline: 1. run the Redis database >> redis-server 2. migrate the files to Redis >> ruby redis_adapter.rb - currently hard-coded to work on "graph-07-04-2011" and "graph-07-11-2011", must be changed inline for other snapshots 3. run the extractor code >> ./extractor.py "snapshot1_name" "snapshot2_name" "output_prefix" [-v] - note that this expects file names as strings, not the file itself. Omit the ".txt" - output_name is the prefix you want to give your output files. The extractor appends _[train/test]_[s/p] to distinguish between training and testing sets, and between SVM and perceptron format - this takes a long time, and the optional -v flag prints the number of edges extracted 4a. run the analysis code for libSVM >> ./analyze_libsvm.py "training_filename" "test_filename" "output_filename" - again, this expects file names as strings, and not the file itself as an argument - the output files from analyze_libsvm are of the format used for libsvm here 4b. run the analysis code for perceptron >> python perceptron_test.py - may require you to first go to the python-perceptron folder, as well as moving the training and testing files to the python-perceptron/data folder 5. run the evalation file >> ./eval.py < data - "data" is the output file of either the perceptron or libSVM code, and here is the actual file
About
semester project for cs161 spring 2012 - google+ link prediction
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published
Languages
- C++ 41.2%
- Python 29.4%
- C 14.8%
- Ruby 13.9%
- Makefile 0.7%