vowpalwabbit 0.1.0 → 0.1.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: fd6a68e5d6103be4f6f58bd3fb6cb9785bd47382d7787b8d0ebc2c0336d87af3
-  data.tar.gz: 2f290dfdfe0539b4daf3e73dfbb39d04fa97e2f5c9aec08d43b0ee81db1d49b2
+  metadata.gz: fa89f06a4117a04e64013f332f75c1ebfd99542c508934430fd511b824fded94
+  data.tar.gz: 00d4b988b39f03ad1a5625e8eccfb427e96ef9488c964e50c7e436c8e2e4e57b
 SHA512:
-  metadata.gz: 7e63904cf470666bbec00e9096e1701c1824af4b7790aa9d137a8a779eefb6d6833aadf4043f7d941e2f8395dfda532349bada33ffaa5cfa8461a593b9e80088
-  data.tar.gz: 0d6b47d922da2cfefb525f59e718e6f3be4c7e99033647f0f8814b9efde0095399b1b87c5922143cefeb8f0b7207ca4811b8c1c29505ea7b5af8cdc5c9910acc
+  metadata.gz: 37051b8953fddd8f88bc22d2adf66003acfaf8a14beacd4314641616c2e607d5a8c59f572675e94f40759fe142ec491bf6d360493e93f910836ec9db84e909a5
+  data.tar.gz: 2233f54d84f06b8176cf4d37c5cd9387eeaaee06cdfdbd1402bad3fa0117e5f3d1f7df60e453aca8ac17dd5de3ececf1c6db026fb112f9eecf080ecc41e11bb3

data/CHANGELOG.md

@@ -1,3 +1,8 @@
+## 0.1.1
+
+- Added support for compressed files
+- Added support for Numo
+
 ## 0.1.0
 
 - First release

data/README.md

@@ -4,9 +4,11 @@
 
 :fire: Uses the C API for blazing performance
 
+[![Build Status](https://travis-ci.org/ankane/vowpalwabbit.svg?branch=master)](https://travis-ci.org/ankane/vowpalwabbit)
+
 ## Installation
 
-First, [install Vowpal Wabbit](https://vowpalwabbit.org/start.html). For Homebrew, use:
+First, install the [Vowpal Wabbit C++ library](https://vowpalwabbit.org/start.html). For Homebrew, use:
 
 ```sh
 brew install vowpal-wabbit
@@ -30,7 +32,7 @@ y = [1, 2, 3, 4]
 Train a model
 
 ```ruby
-model = VowpalWabbit::Regressor.new(l: 100)
+model = VowpalWabbit::Regressor.new(learning_rate: 100)
 model.fit(x, y)
 ```
 
@@ -91,6 +93,12 @@ Data can be an array of arrays
 [[1, 2, 3], [4, 5, 6]]
 ```
 
+Or a Numo NArray
+
+```ruby
+Numo::DFloat.new(3, 2).seq
+```
+
 Or an array of strings
 
 ```ruby
@@ -105,10 +113,17 @@ Or a path to a file
 
 ```ruby
 model.fit("train.txt")
+model.partial_fit("train.txt")
 model.predict("train.txt")
 model.score("train.txt")
 ```
 
+Files can be compressed
+
+```ruby
+model.fit("train.txt.gz")
+```
+
 Read more about the [input format](https://github.com/VowpalWabbit/vowpal_wabbit/wiki/Input-format)
 
 ## History

data/lib/vowpalwabbit/model.rb

@@ -51,7 +51,7 @@ module VowpalWabbit
       bin_str = File.binread(filename)
       model_data = ::FFI::MemoryPointer.new(:char, bin_str.bytesize)
       model_data.put_bytes(0, bin_str)
-      @handle = FFI.VW_InitializeWithModel(param_str, model_data, bin_str.bytesize)
+      @handle = FFI.VW_InitializeWithModel(param_str(@params), model_data, bin_str.bytesize)
       nil
     end
 
@@ -59,12 +59,12 @@ module VowpalWabbit
 
     # TODO clean-up handle
     def handle
-      @handle ||= FFI.VW_InitializeA(param_str)
+      @handle ||= FFI.VW_InitializeA(param_str(@params))
     end
 
-    def param_str
+    def param_str(params)
       args =
-        @params.map do |k, v|
+        params.map do |k, v|
           check_param(k.to_s)
           check_param(v.to_s)
 
@@ -108,31 +108,39 @@ module VowpalWabbit
       end
     end
 
-    # TODO support compressed files
     def each_example(x, y = nil)
-      each_line(x, y) do |line|
-        example = FFI.VW_ReadExampleA(handle, line)
-        yield example
-        FFI.VW_FinishExample(handle, example)
-      end
-    end
-
-    def each_line(x, y)
+      # path to a file
       if x.is_a?(String)
         raise ArgumentError, "Cannot pass y with file" if y
 
-        File.foreach(x) do |line|
-          yield line
+        file_handle = FFI.VW_InitializeA(param_str(data: x, quiet: true))
+        FFI.VW_StartParser(file_handle)
+        loop do
+          example = FFI.VW_GetExample(file_handle)
+          break if example.read_pointer.null?
+          yield example
+          FFI.VW_FinishExample(file_handle, example)
         end
+        FFI.VW_EndParser(file_handle)
+        FFI.VW_Finish(file_handle)
       else
-        raise ArgumentError, "x and y must have same size" if y && x.size != y.size
+        x = x.to_a
+        if y
+          y = y.to_a
+          raise ArgumentError, "x and y must have same size" if x.size != y.size
+        end
 
         x.zip(y || []) do |xi, yi|
-          if xi.is_a?(String)
-            yield xi
-          else
-            yield "#{yi} 1 | #{xi.map.with_index { |v, i| "#{i}:#{v}" }.join(" ")}"
-          end
+          line =
+            if xi.is_a?(String)
+              xi
+            else
+              "#{yi} 1 | #{xi.map.with_index { |v, i| "#{i}:#{v}" }.join(" ")}"
+            end
+
+          example = FFI.VW_ReadExampleA(handle, line)
+          yield example
+          FFI.VW_FinishExample(handle, example)
         end
       end
     end

data/lib/vowpalwabbit/version.rb

@@ -1,3 +1,3 @@
 module VowpalWabbit
-  VERSION = "0.1.0"
+  VERSION = "0.1.1"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: vowpalwabbit
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.1
 platform: ruby
 authors:
 - Andrew Kane
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2019-10-21 00:00:00.000000000 Z
+date: 2019-11-02 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: ffi
@@ -66,6 +66,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '5'
+- !ruby/object:Gem::Dependency
+  name: numo-narray
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 description: 
 email: andrew@chartkick.com
 executables: []