polars-df 0.3.0-x86_64-darwin → 0.4.0-x86_64-darwin

data/lib/polars/plot.rb

@@ -0,0 +1,109 @@
+module Polars
+  module Plot
+    # Plot data.
+    #
+    # @return [Vega::LiteChart]
+    def plot(x = nil, y = nil, type: nil, group: nil, stacked: nil)
+      require "vega"
+
+      raise ArgumentError, "Must specify columns" if columns.size != 2 && (!x || !y)
+      x ||= columns[0]
+      y ||= columns[1]
+      type ||= begin
+        if self[x].numeric? && self[y].numeric?
+          "scatter"
+        elsif self[x].utf8? && self[y].numeric?
+          "column"
+        elsif (self[x].dtype == Date || self[x].dtype.is_a?(Datetime)) && self[y].numeric?
+          "line"
+        else
+          raise "Cannot determine type. Use the type option."
+        end
+      end
+      df = self[(group.nil? ? [x, y] : [x, y, group]).map(&:to_s).uniq]
+      data = df.rows(named: true)
+
+      case type
+      when "line", "area"
+        x_type =
+          if df[x].numeric?
+            "quantitative"
+          elsif df[x].datelike?
+            "temporal"
+          else
+            "nominal"
+          end
+
+        scale = x_type == "temporal" ? {type: "utc"} : {}
+        encoding = {
+          x: {field: x, type: x_type, scale: scale},
+          y: {field: y, type: "quantitative"}
+        }
+        encoding[:color] = {field: group} if group
+
+        Vega.lite
+          .data(data)
+          .mark(type: type, tooltip: true, interpolate: "cardinal", point: {size: 60})
+          .encoding(encoding)
+          .config(axis: {labelFontSize: 12})
+      when "pie"
+        raise ArgumentError, "Cannot use group option with pie chart" unless group.nil?
+
+        Vega.lite
+          .data(data)
+          .mark(type: "arc", tooltip: true)
+          .encoding(
+            color: {field: x, type: "nominal", sort: "none", axis: {title: nil}, legend: {labelFontSize: 12}},
+            theta: {field: y, type: "quantitative"}
+          )
+          .view(stroke: nil)
+      when "column"
+        encoding = {
+          x: {field: x, type: "nominal", sort: "none", axis: {labelAngle: 0}},
+          y: {field: y, type: "quantitative"}
+        }
+        if group
+          encoding[:color] = {field: group}
+          encoding[:xOffset] = {field: group} unless stacked
+        end
+
+        Vega.lite
+          .data(data)
+          .mark(type: "bar", tooltip: true)
+          .encoding(encoding)
+          .config(axis: {labelFontSize: 12})
+      when "bar"
+        encoding = {
+          # TODO determine label angle
+          y: {field: x, type: "nominal", sort: "none", axis: {labelAngle: 0}},
+          x: {field: y, type: "quantitative"}
+        }
+        if group
+          encoding[:color] = {field: group}
+          encoding[:yOffset] = {field: group} unless stacked
+        end
+
+        Vega.lite
+          .data(data)
+          .mark(type: "bar", tooltip: true)
+          .encoding(encoding)
+          .config(axis: {labelFontSize: 12})
+      when "scatter"
+        encoding = {
+          x: {field: x, type: "quantitative", scale: {zero: false}},
+          y: {field: y, type: "quantitative", scale: {zero: false}},
+          size: {value: 60}
+        }
+        encoding[:color] = {field: group} if group
+
+        Vega.lite
+          .data(data)
+          .mark(type: "circle", tooltip: true)
+          .encoding(encoding)
+          .config(axis: {labelFontSize: 12})
+      else
+        raise ArgumentError, "Invalid type: #{type}"
+      end
+    end
+  end
+end

data/lib/polars/series.rb

@@ -1776,8 +1776,9 @@ module Polars
     #   s.is_datelike
     #   # => true
     def is_datelike
-      [Date, Datetime, Duration, Time].include?(dtype)
+      [Date, Time].include?(dtype) || dtype.is_a?(Datetime) || dtype.is_a?(Duration)
     end
+    alias_method :datelike?, :is_datelike
 
     # Check if this Series has floating point numbers.
     #
@@ -1823,8 +1824,45 @@ module Polars
     # def view
     # end
 
-    # def to_numo
-    # end
+    # Convert this Series to a Numo array. This operation clones data but is completely safe.
+    #
+    # @return [Numo::NArray]
+    #
+    # @example
+    #   s = Polars::Series.new("a", [1, 2, 3])
+    #   s.to_numo
+    #   # =>
+    #   # Numo::Int64#shape=[3]
+    #   # [1, 2, 3]
+    def to_numo
+      if !has_validity
+        if is_datelike
+          Numo::RObject.cast(to_a)
+        elsif is_numeric
+          # TODO make more efficient
+          {
+            UInt8 => Numo::UInt8,
+            UInt16 => Numo::UInt16,
+            UInt32 => Numo::UInt32,
+            UInt64 => Numo::UInt64,
+            Int8 => Numo::Int8,
+            Int16 => Numo::Int16,
+            Int32 => Numo::Int32,
+            Int64 => Numo::Int64,
+            Float32 => Numo::SFloat,
+            Float64 => Numo::DFloat
+          }.fetch(dtype).cast(to_a)
+        elsif is_boolean
+          Numo::Bit.cast(to_a)
+        else
+          _s.to_numo
+        end
+      elsif is_datelike
+        Numo::RObject.cast(to_a)
+      else
+        _s.to_numo
+      end
+    end
 
     # Set masked values.
     #
@@ -3493,6 +3531,13 @@ module Polars
       ListNameSpace.new(self)
     end
 
+    # Create an object namespace of all binary related methods.
+    #
+    # @return [BinaryNameSpace]
+    def bin
+      BinaryNameSpace.new(self)
+    end
+
     # Create an object namespace of all categorical related methods.
     #
     # @return [CatNameSpace]
@@ -3757,7 +3802,8 @@ module Polars
       UInt32 => RbSeries.method(:new_opt_u32),
       UInt64 => RbSeries.method(:new_opt_u64),
       Boolean => RbSeries.method(:new_opt_bool),
-      Utf8 => RbSeries.method(:new_str)
+      Utf8 => RbSeries.method(:new_str),
+      Binary => RbSeries.method(:new_binary)
     }
 
     SYM_TYPE_TO_CONSTRUCTOR = {

data/lib/polars/string_expr.rb

@@ -332,7 +332,7 @@ module Polars
     #   # │ -0001   │
     #   # │ 00000   │
     #   # │ 00001   │
-    #   # │ ...     │
+    #   # │ …       │
     #   # │ 10000   │
     #   # │ 100000  │
     #   # │ 1000000 │

data/lib/polars/utils.rb

@@ -93,8 +93,12 @@ module Polars
       Polars.lit(value)
     end
 
-    def self.format_path(path)
-      File.expand_path(path)
+    def self.normalise_filepath(path, check_not_directory: true)
+      path = File.expand_path(path)
+      if check_not_directory && File.exist?(path) && Dir.exist?(path)
+        raise ArgumentError, "Expected a file path; #{path} is a directory"
+      end
+      path
     end
 
     # TODO fix
@@ -216,5 +220,9 @@ module Polars
         val.is_a?(Array) && _is_iterable_of(val, String)
       end
     end
+
+    def self.local_file?(file)
+      Dir.glob(file).any?
+    end
   end
 end

data/lib/polars/version.rb

@@ -1,4 +1,4 @@
 module Polars
   # @private
-  VERSION = "0.3.0"
+  VERSION = "0.4.0"
 end

data/lib/polars.rb

@@ -12,9 +12,12 @@ require "stringio"
 # modules
 require_relative "polars/expr_dispatch"
 require_relative "polars/batched_csv_reader"
+require_relative "polars/binary_expr"
+require_relative "polars/binary_name_space"
 require_relative "polars/cat_expr"
 require_relative "polars/cat_name_space"
 require_relative "polars/convert"
+require_relative "polars/plot"
 require_relative "polars/data_frame"
 require_relative "polars/data_types"
 require_relative "polars/date_time_expr"

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: polars-df
 version: !ruby/object:Gem::Version
-  version: 0.3.0
+  version: 0.4.0
 platform: x86_64-darwin
 authors:
 - Andrew Kane
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2023-02-16 00:00:00.000000000 Z
+date: 2023-04-01 00:00:00.000000000 Z
 dependencies: []
 description: 
 email: andrew@ankane.org
@@ -29,6 +29,8 @@ files:
 - lib/polars/3.1/polars.bundle
 - lib/polars/3.2/polars.bundle
 - lib/polars/batched_csv_reader.rb
+- lib/polars/binary_expr.rb
+- lib/polars/binary_name_space.rb
 - lib/polars/cat_expr.rb
 - lib/polars/cat_name_space.rb
 - lib/polars/convert.rb
@@ -49,6 +51,7 @@ files:
 - lib/polars/list_expr.rb
 - lib/polars/list_name_space.rb
 - lib/polars/meta_expr.rb
+- lib/polars/plot.rb
 - lib/polars/rolling_group_by.rb
 - lib/polars/series.rb
 - lib/polars/slice.rb