carray-dataframe 1.0.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: 7f93c348e3fd8e166ddba89d5ddc1f0fb86653a6
+  data.tar.gz: cb16a824a9e0c2aaf40db3f26328176ef9eb882e
+SHA512:
+  metadata.gz: 8f96c5cf91470a732a09aa1e3629b94963d29d786dbfa9768430c18e0ebe1b2219f7256ddc17c783ce8cac9253c19c4cbce9e16d005435f0f718f97c788b832d
+  data.tar.gz: b982e8a4b8f162f69bb86fd31d0d92d1d86c7c06b663b9cc091974fe46fcdbf16635e3e793cc91b813fd54cff29e7135416c449ba0089985332dad354d2a18a9

data/API.txt

@@ -0,0 +1,83 @@
+
+### Constructor
+
+CADataFrame.new(columns_or_table, row_index: nil, column_names: nil) 
+CADataFrame.new(columns_or_table, row_index: nil, column_names: nil) { ... }
+
+  If block specified, arrange is called internaly with the block.
+
+### Attributes
+
+CADataFrame#column_number
+CADataFrame#column_names
+CADataFrame#column_types
+CADataFrame#columns
+CADataFrame#row_index
+CADataFrame#row_number
+
+### Index Access
+
+df[["AAA"]]        => CADataFrame include column "AAA"
+df[["AAA","BBB"]]  => CADataFrame include column "AAA", "BBB"
+
+df["AAA"]               => 1 dimensional CArray 
+df[["AAA"]].to_ca       => 2 dimensional CArray with column_names
+df[["AAA","BBB"]].to_ca => 2 dimensional CArray with column_names
+
+
+df[dfmask] 
+  return new detached CADataFrame masked where dfmask's value equal 1
+
+
+### Iterators
+
+CADataFrame#each_column_name { |name| ... }
+CADataFrame#each_column { |name, column| ... }
+CADataFrame#each_row(with: [Array|Hash]) { |row| ... }
+CADataFrame#each_row_with_row_index(with: [Array|Hash]) { |row, idx| ... }
+
+### Transformation
+
+CADataFrame#add_suffix(suffix_string) -> CADataFrame
+  Add suffix_string to all column names
+
+CADataFrame#transpose(column_names: )
+
+
+### Conversion
+
+CADataFrame#ca      -> CADFArray (Reference Array)
+CADataFrame#to_ca   -> CArray with CA::TableMethods
+CADataFrame#to_hash -> Hash
+CADataFrame#to_xlsx(with_row_index: false) -> Hash
+  Masked element converted to "=NA()"
+
+CADataFrame#columns_to_hash(key_name, *value_names)
+
+ex) df.columns_to_hash("bbb",["aaa","ccc"])
+
+  ---------------
+   aaa  bbb  ccc 
+  ---------------
+   4    10   100 
+   5    20   50  
+   6    30   -30 
+   7    40   -50 
+  ---------------
+
+  {10=>[4, 100], 20=>[5, 50], 30=>[6, -30], 40=>[7, -50]}
+
+
+
+
+CADataFrame 
+ 
+#append(name) { INSTANCE_CONTEXT }    <- any carray
+#lead(name) { INSTANCE_CONTEXT }      <- any carray
+#execute { INSTANCE_CONTEXT }         => any object
+#select(name...) { INSTANCE_CONTEXT } <- boolean carray
+#reorder { INSTANCE_CONTEXT }         <- int32 carray (addresses for mapping)
+#order_by { INSTANCE_CONTEXT }        <- Array of int32 carray or carray (addresses for mapping)
+
+#calculate {|label, column| CALLER_CONTEXT } <- scalar
+#resample {|label, column| CALLER_CONTEXT }  <- any carray

data/README.md

@@ -0,0 +1,5 @@
+carray-dataframe
+================
+
+
+gem install rsruby

data/carray-dataframe.gemspec

@@ -0,0 +1,25 @@
+
+Gem::Specification::new do |s|
+  version = "1.0.0"
+
+  files = Dir.glob("**/*") - [ 
+                               Dir.glob("carray*.gem"), 
+                             ].flatten
+
+  s.platform    = Gem::Platform::RUBY
+  s.name        = "carray-dataframe"
+  s.summary     = "Extension for realizing DataFrame of R in Ruby"
+  s.description = <<-HERE
+    Extension for realizing DataFrame of R in Ruby
+  HERE
+  s.version     = version
+  s.author      = "Hiroki Motoyoshi"
+  s.email       = ""
+  s.homepage    = 'https://github.com/himotoyoshi/carray-dataframe'
+  s.files       = files
+  s.has_rdoc    = false
+  s.required_ruby_version = ">= 1.8.1"
+  s.add_runtime_dependency 'carray', '~> 1.1'
+  s.add_runtime_dependency 'axlsx', '~> 2.0'
+  s.add_runtime_dependency 'spreadsheet', '~> 1.1'
+end

data/examples/R/fit.rb

@@ -0,0 +1,24 @@
+require "carray"
+require "R"
+
+R.run
+
+x = CArray.float(200).span(0..4r)
+v = x.random(4)-2
+
+a = 3
+b = 5
+c = 7
+y = a*x**2 + b*x + c + v
+
+res = R %{ 
+  nls(y ~ a*x^2 + b*x + c, start=c(a=100,b=1,c=1), trace=TRUE) 
+}, :x=>x, :y=>y
+
+a1,b1,c1 = R.coef(res).to_ruby.values_at("a","b","c")
+
+CA.gnuplot {
+  plot [x,y],
+       [x,a1*x**2+b1*x+c, nil, "lines"]
+}
+

data/examples/R/iris.rb

@@ -0,0 +1,9 @@
+require "R"
+
+R.run
+
+iris = R.iris
+
+CA.gnuplot {
+  plot [iris.Sepal_Length, iris.Sepal_Width]
+}

data/examples/R/japan_area.rb

@@ -0,0 +1,30 @@
+#
+# From https://oku.edu.mie-u.ac.jp/~okumura/stat/100410a.html
+#
+
+require "R"
+
+R.run
+
+areaname = ["北海道","本州","四国","九州","沖縄"].to_ca
+areasize = [83457,231113,18792,42191,2276].to_ca / 10000.0
+
+R %{
+  par(family="HiraKakuProN-W3")
+  par(las=1)
+  par(mgp=c(2,0.8,0))
+  barplot(areasize, names.arg=areaname)
+  axis(2, labels="面積 (万km^2)", at=20, hadj=0.3, padj=-1, tick=FALSE) 
+}, :areasize=>areasize, :areaname=>areaname
+
+gets
+
+R {
+  par :family=>"HiraKakuProN-W3"
+  par :las=>1
+  par :mgp=>[2,0.8,0]
+  barplot areasize, "names.arg"=>areaname
+  axis 2, :labels=>"面積 (万km^2)", :at=>20, :hadj=>0.3, :padj=>-1, :tick=>false 
+}
+
+gets

data/examples/R/kyaku.rb

@@ -0,0 +1,22 @@
+require "carray"
+
+CA.gnuplot {
+  terminal %{ wxt }
+  (1..10).each do |n|
+    x = CArray.double(1000000) {0}
+    n.times do 
+      x += CArray.double(1000000).random
+    end
+    x = x/n
+    df = CADataFrame.new(:x=>x)
+    h = df.histogram(:x, CA_DOUBLE(0..1,0.01))
+
+    plot [h.x, h.count, nil, "boxes fill solid 0.5 noborder"], 
+         :x=>[nil, 0..1],
+         :title=>n.to_s,
+         :nopause=>true
+
+    sleep 0.5
+  end
+  gets
+}

data/examples/group_by.rb

@@ -0,0 +1,78 @@
+require "carray"
+
+csv =<<HERE
+name,v1,v2
+A,1,3
+B,3,2
+C,2,1
+B,1,3
+C,1,4
+A,4,2
+B,5,3
+C,3,3
+C,1,1
+C,6,3
+C,8,1
+A,1,2
+HERE
+
+f = CADataFrame.from_csv(csv) {
+  header
+  body
+}.arrange {
+  int :v1, :v2
+}
+
+p f.resample { |l, c|
+  c = c.reshape(false,2)
+  case l
+  when "name"
+    c[nil,-1]
+  else
+    c.max(1)
+  end
+}
+
+p df = CADataFrame.concat(f.calculate(:sum),
+                          f.calculate(:mean)).arrange {
+  eliminate :name
+  append :sum, v1 + v2
+}
+
+p f.group_by(:name).table { 
+  {
+    :count  => row_number, 
+    :v1_sum => v1.sum,
+    :v1_mean => v1.mean,
+    :v2_sum => v2.sum,
+    :v2_mean => v2.mean,
+  }
+}
+
+p f.group_by(:v2).table { 
+  { 
+    :count    => row_number, 
+    :namelist => name.sort.join(""),
+  }
+}
+
+p "--- Pivot"
+p f.pivot({:v1=>CA_INT(1..8)},{:v2=>CA_INT(1..5)}).table {
+  name.size > 0 ? name.join("") : "-"
+}
+
+g = f.group_by(:v1,:v2)
+
+p t = g.table {
+  { 
+    :count    => row_number, 
+    :namelist => name.sort.join(""),
+  }
+}
+
+p g[[1,3]]
+
+p t.select { count >= 2 }
+
+
+

data/examples/hist.rb

@@ -0,0 +1,27 @@
+require "carray"
+text = <<EOS
+name,NAME,a,b,c
+u,U,1,2,3
+v,V,2,3,4
+w,W,5,1,3
+x,X,4,3,1
+y,Y,1,1,2
+z,Z,2,3,1
+EOS
+
+df = CADataFrame.from_csv(text) {
+  header
+  body
+}.arrange {
+  int :a,:b,:c
+}
+
+CA.gnuplot {
+  set %{ style histogram rowstacked }
+  set %{ style fill solid border -1 }
+  plot [df.a, "a", "histogram"],
+       [df.b, "b", "histogram"],
+       [df.c, "c", "histogram"],
+       :x=>["NAME",nil,nil, histogram_tics(df.NAME)],
+       :y=>["VALUE",0..20]
+}

data/examples/iris.rb

@@ -0,0 +1,29 @@
+require "carray"
+require "R"
+require_relative "../lib/carray/dataframe/dataframe"
+
+R.run
+
+df = R.iris
+df.lead "id", df.row_index
+
+p df
+
+df.to_xlsx('out.xlsx')
+
+petal = df[["id", "Petal.Length","Petal.Width"]]
+sepal = df[["id", "Sepal.Length","Sepal.Width"]]
+species = df[["id", "Species"]]
+
+
+p df["Species"].value_counts
+
+#p d2 = df.to_daru
+
+tbl = df.to_sql("iris").to_df %{
+  select * from iris order by Sepal_Width desc;
+}
+
+p tbl
+
+#tbl.to_xlsx("out.xlsx")

data/examples/map.rb

@@ -0,0 +1,23 @@
+require "carray"
+
+csv =<<CSV
+id,gender,ans1,ans2
+1,F,1,0
+2,F,0,0
+3,M,1,0
+4,M,0,1
+5,F,1,1
+CSV
+
+df = CADataFrame.from_csv(csv) {
+  header
+  body
+}.arrange {
+  int :id,:ans1,:ans2
+  append :ans1s, ["NG","OK"].to_ca[ans1]
+  append :ans2s, ["NG","OK"].values_at(*ans2.to_a)
+}
+
+p df
+
+#p df.group_by(:gender).calculate(:sum)

data/examples/match.rb

@@ -0,0 +1,21 @@
+require "carray"
+
+df = CADataFrame.new a: [1,2,3,5,6,7,9,10],
+                     b: [30,20,20,30,20,10,20,30],
+                     c: [2,1,1,1,2,1,2,2]
+
+a1 = CA_INT([1,2,5,6,7,9])
+a2 = CA_INT([1,2,3,4,5,6,7,8,10,11])
+
+p df
+p df.matchup(:a, a1)
+df2 = df.matchup(:a, a2)
+
+df2.arrange {
+  unmask -9999, :b, :c
+}
+
+p df2
+
+b1 = CA_INT([10,20,30])
+p df.matchup(:b, b1)

data/examples/test1.rb

@@ -0,0 +1,44 @@
+require "carray"
+require_relative "../lib/carray/dataframe/dataframe"
+
+hash =   { 'aaa' => [4,5,6,7], 'bbb' => [10,20,30,40], 'ccc' => [100,50,-30,-50] }
+p df = CADataFrame.new(hash)
+
+table = CA_OBJECT([[4,5,6,7], [10,20,30,40], [100,50,-30,-50]]).t
+p df = CADataFrame.new(table, column_names: ['aaa','bbb','ccc'])
+
+table.extend(CA::TableMethods)
+table.column_names = ['aaa','bbb','ccc']
+p df = CADataFrame.new(table)
+
+hash =   { 'aaa' => [4,5,6,7], 'bbb' => [10,20,30,40], 'ccc' => [100,50,-30,-50] }
+p df = CADataFrame.new(hash, row_index: ["a","b","c","d"])
+
+df.each_row(with: Array) {|row|
+  p row
+}
+
+df.each_row_with_row_index(with: Array) {|row,i|
+  p [row,i]
+}
+
+p df.to_ca.column_names
+p df.to_hash
+p df.columns_to_hash("bbb","aaa")
+p df.columns_to_hash("bbb",["aaa"])
+p df.columns_to_hash("bbb",["aaa","ccc"])
+
+p df.add_suffix("_no")
+p df.transpose
+
+p df2 = df.to_df
+
+p df["aaa"]
+p b = df[["aaa"]]
+#b.detouch!
+
+b[0,0] = -1111
+
+p df
+p df2
+