daru_lite 0.1

data/lib/daru_lite/version.rb

@@ -0,0 +1,3 @@
+module DaruLite
+  VERSION = '0.1'.freeze
+end

data/lib/daru_lite.rb

@@ -0,0 +1,99 @@
+# :nocov:
+def jruby?
+  RUBY_ENGINE == 'jruby'
+end
+# :nocov:
+
+module DaruLite
+  DAYS_OF_WEEK = {
+    'SUN' => 0,
+    'MON' => 1,
+    'TUE' => 2,
+    'WED' => 3,
+    'THU' => 4,
+    'FRI' => 5,
+    'SAT' => 6
+  }.freeze
+
+  MONTH_DAYS = {
+    1 => 31,
+    2 => 28,
+    3 => 31,
+    4 => 30,
+    5 => 31,
+    6 => 30,
+    7 => 31,
+    8 => 31,
+    9 => 30,
+    10 => 31,
+    11 => 30,
+    12 => 31
+  }.freeze
+
+  MISSING_VALUES = [nil, Float::NAN].freeze
+
+  @lazy_update = false
+
+  SPLIT_TOKEN = ','.freeze
+
+  @error_stream = $stderr
+
+  class << self
+    # A variable which will set whether Vector metadata is updated immediately or lazily.
+    # Call the #update method every time a values are set or removed in order to update
+    # metadata like positions of missing values.
+    attr_accessor :lazy_update, :error_stream
+
+    def create_has_library(library)
+      lib_underscore = library.to_s.tr('-', '_')
+      define_singleton_method(:"has_#{lib_underscore}?") do
+        cv = "@@#{lib_underscore}"
+        unless class_variable_defined? cv
+          begin
+            require library.to_s
+            class_variable_set(cv, true)
+          rescue LoadError
+            # :nocov:
+            class_variable_set(cv, false)
+            # :nocov:
+          end
+        end
+        class_variable_get(cv)
+      end
+    end
+
+    def error(msg)
+      error_stream&.puts msg
+    end
+  end
+end
+
+autoload :CSV, 'csv'
+require 'matrix'
+require 'forwardable'
+require 'erb'
+require 'date'
+
+require 'daru_lite/version'
+
+require 'open-uri'
+
+require 'daru_lite/index/index'
+require 'daru_lite/index/multi_index'
+require 'daru_lite/index/categorical_index'
+
+require 'daru_lite/helpers/array'
+require 'daru_lite/configuration'
+require 'daru_lite/vector'
+require 'daru_lite/dataframe'
+require 'daru_lite/monkeys'
+require 'daru_lite/formatters/table'
+require 'daru_lite/iruby/helpers'
+require 'daru_lite/exceptions'
+
+require 'daru_lite/core/group_by'
+require 'daru_lite/core/query'
+require 'daru_lite/core/merge'
+
+require 'daru_lite/date_time/offsets'
+require 'daru_lite/date_time/index'

data/profile/_base.rb

@@ -0,0 +1,23 @@
+$:.unshift File.expand_path("../../lib", __FILE__)
+
+require 'ruby-prof'
+require 'fileutils'
+
+require 'daru_lite'
+
+def __profile__(name = nil)
+  # infers name to be "sorting" when called from "profile/sorting.rb:10:in `<main>'"
+  name ||= caller.first.split(':').first.split('/').last.sub('.rb', '')
+
+  path = File.expand_path("../out/#{name}.html", __FILE__)
+  FileUtils.mkdir_p File.dirname(path)
+
+  RubyProf.start
+
+  yield
+
+  res = RubyProf.stop
+  RubyProf::GraphHtmlPrinter.new(res)
+    .print(File.open(path, 'w'))
+
+end

data/profile/df_to_a.rb

@@ -0,0 +1,10 @@
+require_relative '_base'
+
+n = 40_000
+keys = (1..(n)).to_a
+
+df = DaruLite::DataFrame.new(idx: 1.upto(n).to_a, keys: 1.upto(n).map { |v| keys[Random.rand(n)]})
+
+__profile__ do
+  df.to_a
+end

data/profile/filter.rb

@@ -0,0 +1,13 @@
+require_relative '_base'
+
+df = DaruLite::DataFrame.new({
+  a: [1,2,3,4,5,6]*1000,
+  b: ['a','b','c','d','e','f']*1000,
+  c: [11,22,33,44,55,66]*1000
+}, index: (1..6000).to_a.shuffle)
+
+__profile__ do
+  df.filter(:row) do |r|
+    r[:a] == 2 or r[:c] == 55
+  end
+end

data/profile/joining.rb

@@ -0,0 +1,13 @@
+require_relative '_base'
+
+n = 40_000
+keys = (1..(n)).to_a
+base_data = { idx: 1.upto(n).to_a, keys: 1.upto(n).map { |v| keys[Random.rand(n)]}}
+lookup_hash = keys.map { |k| [k, k * 100]}.to_h
+
+base_data_df = DaruLite::DataFrame.new(base_data)
+lookup_df = DaruLite::DataFrame.new(keys: lookup_hash.keys, values: lookup_hash.values)
+
+__profile__ do
+  base_data_df.join(lookup_df, on: [:keys], how: :inner)
+end

data/profile/sorting.rb

@@ -0,0 +1,12 @@
+require_relative '_base'
+
+vector = DaruLite::Vector.new(10_000.times.map.to_a.shuffle)
+df = DaruLite::DataFrame.new({
+  a: vector,
+  b: vector,
+  c: vector
+})
+
+__profile__ do
+  df.sort([:a])
+end

data/profile/vector_each_with_index.rb

@@ -0,0 +1,9 @@
+require_relative '_base'
+
+vector = DaruLite::Vector.new(['a','b','c','d','e','f']*1000, index: (1..6000).to_a.shuffle)
+
+__profile__ do
+  100.times do
+    vector.each_with_index{|val, i| }
+  end
+end

data/profile/vector_new.rb

@@ -0,0 +1,9 @@
+require_relative '_base'
+
+n = 40_0000
+idx = (1..n).to_a.map(&:to_s)
+
+
+__profile__ do
+  DaruLite::Vector.new(1..n, index: idx)
+end

data/spec/accessors/array_wrapper_spec.rb

@@ -0,0 +1,3 @@
+describe DaruLite::Accessors::ArrayWrapper do
+  # TODO:
+end