README.md in libmf-0.1.3 vs README.md in libmf-0.2.0
- old
+ new
@@ -1,11 +1,13 @@
# LIBMF
[LIBMF](https://github.com/cjlin1/libmf) - large-scale sparse matrix factorization - for Ruby
-[](https://travis-ci.org/ankane/libmf)
+Check out [Disco](https://github.com/ankane/disco) for higher-level collaborative filtering
+[](https://travis-ci.org/ankane/libmf) [](https://ci.appveyor.com/project/ankane/libmf/branch/master)
+
## Installation
Add this line to your application’s Gemfile:
```ruby
@@ -94,12 +96,11 @@
lambda_q2: 0.1, # coefficient of L2-norm regularization on Q
learning_rate: 0.1, # learning rate
alpha: 0.1, # importance of negative entries
c: 0.0001, # desired value of negative entries
nmf: false, # perform non-negative MF (NMF)
- quiet: false, # no outputs to stdout
- copy_data: true # copy data in training procedure
+ quiet: false # no outputs to stdout
)
```
### Loss Functions
@@ -119,10 +120,36 @@
- 10 - row-oriented pair-wise logarithmic loss
- 11 - column-oriented pair-wise logarithmic loss
- 12 - squared error (L2-norm)
+## Performance
+
+For performance, read data directly from files
+
+```ruby
+model.fit("train.txt", eval_set: "validate.txt")
+model.cv("train.txt")
+```
+
+Data should be in the format `row_index column_index value`:
+
+```txt
+0 0 5.0
+0 2 3.5
+1 1 4.0
+```
+
+## Numo
+
+Get latent factors as Numo arrays
+
+```ruby
+model.p_factors(format: :numo)
+model.q_factors(format: :numo)
+```
+
## Resources
- [LIBMF: A Library for Parallel Matrix Factorization in Shared-memory Systems](https://www.csie.ntu.edu.tw/~cjlin/papers/libmf/libmf_open_source.pdf)
## History
@@ -135,5 +162,15 @@
- [Report bugs](https://github.com/ankane/libmf/issues)
- Fix bugs and [submit pull requests](https://github.com/ankane/libmf/pulls)
- Write, clarify, or fix documentation
- Suggest or add new features
+
+To get started with development:
+
+```sh
+git clone --recursive https://github.com/ankane/libmf.git
+cd libmf
+bundle install
+bundle exec rake vendor:all
+bundle exec rake test
+```