breakout-detection 0.1.0 → 0.1.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 8f81b5a2a9781787903f293322eef1ff6d08479cde2df25eea47c3023cd501bd
-  data.tar.gz: 1839625455de87fca184eb2f1fdf8bc7a2e8809c8afd17b9fc9ec61d6d775b7a
+  metadata.gz: 6678a5f3c82af3a88809391a6c692f3d0af7d6d6acb335936a6a0712ed0a2176
+  data.tar.gz: d44d0f92be5412e5f49e095f9b3a1d0c564c30ac2e42911e276bc1092c10edd3
 SHA512:
-  metadata.gz: 1fa14ae2cc9db547bdeb13a3232d87c96e239aa6dde618ea6b7fa9c53aaca5596b45223c3e7eeceb9a8aa0428de77c54c4e8275adf57b9767c13a889538981bb
-  data.tar.gz: 4ba5ee1decd2ab9084c3d377ec6d37a743c7f5ff47bb48eb55bdd5ecb9bce57df534522643c448b00a35cdd39d1a429ebc1c036440b079046c321bbef5e193a1
+  metadata.gz: 343f0d3245d495d022b2953009f724d0bf83f2448280b4e9ad0b4ca678241ac57044b9d6b3efbb2d473f2fc61d0083e601fbae669b243196aba080e440296bce
+  data.tar.gz: 129bc113eae6cda7e081ad8cc15f7b01f772bd7504187b96ad0e2a3af583e256d7127049e1528c5de4c9eeb87dad2e9db00101c35e14def788243f073226c36a

data/CHANGELOG.md

@@ -1,3 +1,7 @@
+## 0.1.1 (2021-10-20)
+
+- Added `plot` method
+
 ## 0.1.0 (2021-09-02)
 
 - First release

data/README.md

@@ -2,7 +2,7 @@
 
 :fire: [BreakoutDetection](https://github.com/twitter/BreakoutDetection) for Ruby
 
-Learn more about [how it works](https://blog.twitter.com/engineering/en_us/a/2014/breakout-detection-in-the-wild)
+Learn [how it works](https://blog.twitter.com/engineering/en_us/a/2014/breakout-detection-in-the-wild)
 
 [![Build Status](https://github.com/ankane/breakout/workflows/build/badge.svg?branch=master)](https://github.com/ankane/breakout/actions)
 
@@ -60,6 +60,20 @@ Breakout.detect(
 )
 ```
 
+## Plotting
+
+Add [Vega](https://github.com/ankane/vega) to your application’s Gemfile:
+
+```ruby
+gem 'vega'
+```
+
+And use:
+
+```ruby
+Breakout.plot(series, breakouts)
+```
+
 ## Credits
 
 This library uses the C++ code from the [BreakoutDetection](https://github.com/twitter/BreakoutDetection) R package and is available under the same license.

data/ext/breakout/edm_multi.cpp

@@ -50,7 +50,7 @@ std::vector<int> EDM_multi(const std::vector<double>& Z, int min_size = 24, doub
     for (int i = min_size - 1; i < s; ++i)
       insert_element(right_min, right_max, Z[i]);
 
-    // Iterate over possible locations for the penultiamte change
+    // Iterate over possible locations for the penultimate change
     for (int t = min_size; t < s - min_size + 1; ++t) { // modify limits to deal with min_size
       insert_element(left_min, left_max, Z[t - 1]);     // insert element into left tree
       remove_element(right_min, right_max, Z[t - 1]);   // remove element from right tree

data/ext/breakout/edm_percent.cpp

@@ -1,7 +1,7 @@
 /*
-  Penalizes based on percent chagne in the statistic value.
-  Linear penalty means that each new breakout must result in an at least X% increast
-  Quadratic penalty means that each new brekaout must result in at least an (X*k)% increase for k breakouts
+  Penalizes based on percent change in the statistic value.
+  Linear penalty means that each new breakout must result in an at least X% increase
+  Quadratic penalty means that each new breakout must result in at least an (X*k)% increase for k breakouts
 */
 
 #include <algorithm>

data/ext/breakout/edm_tail.cpp

@@ -355,7 +355,6 @@ void BackwardUpdate(std::vector<double>& Z, Information& info, int& tau1, double
       index /= 2;
     }
   }
-  double qb = std::pow(GetQuantile(info.B, quant), alpha);
   // Move tau2 from the end of the time series to the front.
   // Update the statistic value along the way
   tau2 = N;
@@ -366,7 +365,7 @@ void BackwardUpdate(std::vector<double>& Z, Information& info, int& tau1, double
       --info.B[index];
       index /= 2;
     }
-    qb = std::pow(GetQuantile(info.B, quant), alpha);
+    double qb = std::pow(GetQuantile(info.B, quant), alpha);
 
     double stat = 2 * qc - qa - qb;
     stat *= (double)(tau2 - tau1) * tau1 / tau2;

data/ext/breakout/edmx.cpp

@@ -1,7 +1,7 @@
 /*
 Robust estimation of 2[mean(X)-mean(Y)]^2 time normalization factor
 This is the E-Divisive E-statistic when alpha = 2
-Instead of calculating mean(X) we calculate median(X), and similarly for Y
+Instead of calculating mean(X), we calculate median(X), and similarly for Y
 */
 
 #include <algorithm>

data/lib/breakout/version.rb

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

data/lib/breakout.rb

@@ -5,24 +5,78 @@ require "breakout/ext"
 require "breakout/version"
 
 module Breakout
-  def self.detect(series, min_size: 30, method: "multi", alpha: 2, beta: nil, degree: 1, percent: nil, exact: true)
-    raise ArgumentError, "min_size must be at least 2" if min_size < 2
-    raise ArgumentError, "beta and percent cannot be passed together" unless beta.nil? || percent.nil?
-    raise ArgumentError, "alpha must be between 0 and 2" if alpha < 0 || alpha > 2
-    raise ArgumentError, "degree must be 0, 1, or 2" unless [0, 1, 2].include?(degree)
-    raise ArgumentError, "Bad method" unless ["amoc", "multi"].include?(method)
-
-    return [] if series.size < min_size
-
-    if series.is_a?(Hash)
-      sorted = series.sort_by { |k, _| k }
-      z = sorted.map(&:last)
-    else
-      z = series
+  class << self
+    def detect(series, min_size: 30, method: "multi", alpha: 2, beta: nil, degree: 1, percent: nil, exact: true)
+      raise ArgumentError, "min_size must be at least 2" if min_size < 2
+      raise ArgumentError, "beta and percent cannot be passed together" unless beta.nil? || percent.nil?
+      raise ArgumentError, "alpha must be between 0 and 2" if alpha < 0 || alpha > 2
+      raise ArgumentError, "degree must be 0, 1, or 2" unless [0, 1, 2].include?(degree)
+      raise ArgumentError, "method must be amoc or multi" unless ["amoc", "multi"].include?(method)
+
+      return [] if series.size < min_size
+
+      if series.is_a?(Hash)
+        sorted = series.sort_by { |k, _| k }
+        z = sorted.map(&:last)
+      else
+        z = series
+      end
+
+      res = _detect(z, min_size, method, alpha, beta, degree, percent, exact)
+      res.map! { |i| sorted[i][0] } if series.is_a?(Hash)
+      res
+    end
+
+    def plot(series, breakouts)
+      require "vega"
+
+      data =
+        if series.is_a?(Hash)
+          series.map { |k, v| {x: iso8601(k), y: v, breakout: breakouts.include?(k)} }
+        else
+          series.map.with_index { |v, i| {x: i, y: v, breakout: breakouts.include?(i)} }
+        end
+
+      if series.is_a?(Hash)
+        x = {field: "x", type: "temporal"}
+        x["scale"] = {type: "utc"} if series.keys.first.is_a?(Date)
+      else
+        x = {field: "x", type: "quantitative"}
+      end
+
+      Vega.lite
+        .data(data)
+        .layer([
+          {
+            mark: {type: "line"},
+            encoding: {
+              x: x,
+              y: {field: "y", type: "quantitative", scale: {zero: false}},
+              color: {value: "#fa9088"}
+            }
+          },
+          {
+            transform: [{"filter": "datum.breakout == true"}],
+            mark: {type: "rule"},
+            encoding: {
+              x: x,
+              color: {value: "#19c7ca"},
+              strokeWidth: {value: 2},
+              strokeDash: {value: [6, 6]}
+            }
+          }
+        ])
+        .config(axis: {title: nil, labelFontSize: 12})
     end
 
-    res = _detect(z, min_size, method, alpha, beta, degree, percent, exact)
-    res.map! { |i| sorted[i][0] } if series.is_a?(Hash)
-    res
+    private
+
+    def iso8601(v)
+      if v.is_a?(Date)
+        v.strftime("%Y-%m-%d")
+      else
+        v.strftime("%Y-%m-%dT%H:%M:%S.%L%z")
+      end
+    end
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: breakout-detection
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.1
 platform: ruby
 authors:
 - Andrew Kane
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2021-09-03 00:00:00.000000000 Z
+date: 2021-10-20 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rice