polars-df 0.1.3 → 0.1.4

data/lib/polars/expr_dispatch.rb

@@ -3,11 +3,19 @@ module Polars
   module ExprDispatch
     private
 
+    def self.included(base)
+      base.attr_accessor :_s
+      base.singleton_class.attr_accessor :_accessor
+    end
+
     def method_missing(method, ...)
       return super unless self.class.method_defined?(method)
 
+      namespace = self.class._accessor
+
       s = Utils.wrap_s(_s)
       expr = Utils.col(s.name)
+      expr = expr.send(namespace) if namespace
       s.to_frame.select(expr.send(method, ...)).to_series
     end
   end

data/lib/polars/group_by.rb

@@ -1,4 +1,5 @@
 module Polars
+  # Starts a new GroupBy operation.
   class GroupBy
     # @private
     attr_accessor :_df, :_dataframe_class, :by, :maintain_order

data/lib/polars/io.rb

@@ -183,7 +183,7 @@ module Polars
     # @param has_header [Boolean]
     #   Indicate if the first row of dataset is a header or not.
     #   If set to false, column names will be autogenerated in the
-    #   following format: ``column_x``, with ``x`` being an
+    #   following format: `column_x`, with `x` being an
     #   enumeration over every column in the dataset starting at 1.
     # @param sep [String]
     #   Single byte character to use as delimiter in the file.

data/lib/polars/lazy_frame.rb

@@ -558,7 +558,7 @@ module Polars
     #       "ham" => ["a", "b", "c"]
     #     }
     #   ).lazy
-    #   lf.filter(Polars.col("foo") < 3).collect()
+    #   lf.filter(Polars.col("foo") < 3).collect
     #   # =>
     #   # shape: (2, 3)
     #   # ┌─────┬─────┬─────┐
@@ -910,7 +910,7 @@ module Polars
     #     [
     #       (Polars.col("a") ** 2).alias("a^2"),
     #       (Polars.col("b") / 2).alias("b/2"),
-    #       (Polars.col("c").is_not()).alias("not c")
+    #       (Polars.col("c").is_not).alias("not c")
     #     ]
     #   ).collect
     #   # =>
@@ -1256,8 +1256,12 @@ module Polars
       select(Utils.col("*").take_every(n))
     end
 
-    # def fill_null
-    # end
+    # Fill null values using the specified value or strategy.
+    #
+    # @return [LazyFrame]
+    def fill_null(value = nil, strategy: nil, limit: nil, matches_supertype: nil)
+      select(Polars.all.fill_null(value, strategy: strategy, limit: limit))
+    end
 
     # Fill floating point NaN values.
     #

data/lib/polars/lazy_functions.rb

@@ -191,8 +191,16 @@ module Polars
       end
     end
 
-    # def n_unique
-    # end
+    # Count unique values.
+    #
+    # @return [Object]
+    def n_unique(column)
+      if column.is_a?(Series)
+        column.n_unique
+      else
+        col(column).n_unique
+      end
+    end
 
     # Get the first value.
     #
@@ -213,14 +221,61 @@ module Polars
       end
     end
 
-    # def last
-    # end
+    # Get the last value.
+    #
+    # Depending on the input type this function does different things:
+    #
+    # - nil -> expression to take last column of a context.
+    # - String -> syntactic sugar for `Polars.col(..).last`
+    # - Series -> Take last value in `Series`
+    #
+    # @return [Object]
+    def last(column = nil)
+      if column.nil?
+        return Utils.wrap_expr(_last)
+      end
 
-    # def head
-    # end
+      if column.is_a?(Series)
+        if column.len > 0
+          return column[-1]
+        else
+          raise IndexError, "The series is empty, so no last value can be returned"
+        end
+      end
+      col(column).last
+    end
 
-    # def tail
-    # end
+    # Get the first `n` rows.
+    #
+    # @param column [Object]
+    #   Column name or Series.
+    # @param n [Integer]
+    #   Number of rows to return.
+    #
+    # @return [Object]
+    def head(column, n = 10)
+      if column.is_a?(Series)
+        column.head(n)
+      else
+        col(column).head(n)
+      end
+    end
+
+    # Get the last `n` rows.
+    #
+    # @param column [Object]
+    #   Column name or Series.
+    # @param n [Integer]
+    #   Number of rows to return.
+    #
+    # @return [Object]
+    def tail(column, n = 10)
+      if column.is_a?(Series)
+        column.tail(n)
+      else
+        col(column).tail(n)
+      end
+    end
 
     # Return an expression representing a literal value.
     #
@@ -242,14 +297,69 @@ module Polars
     # def cumsum
     # end
 
-    # def spearman_rank_corr
-    # end
+    # Compute the spearman rank correlation between two columns.
+    #
+    # Missing data will be excluded from the computation.
+    #
+    # @param a [Object]
+    #   Column name or Expression.
+    # @param b [Object]
+    #   Column name or Expression.
+    # @param ddof [Integer]
+    #   Delta degrees of freedom
+    # @param propagate_nans [Boolean]
+    #   If `True` any `NaN` encountered will lead to `NaN` in the output.
+    #   Defaults to `False` where `NaN` are regarded as larger than any finite number
+    #   and thus lead to the highest rank.
+    #
+    # @return [Expr]
+    def spearman_rank_corr(a, b, ddof: 1, propagate_nans: false)
+      if a.is_a?(String)
+        a = col(a)
+      end
+      if b.is_a?(String)
+        b = col(b)
+      end
+      Utils.wrap_expr(RbExpr.spearman_rank_corr(a._rbexpr, b._rbexpr, ddof, propagate_nans))
+    end
 
-    # def pearson_corr
-    # end
+    # Compute the pearson's correlation between two columns.
+    #
+    # @param a [Object]
+    #   Column name or Expression.
+    # @param b [Object]
+    #   Column name or Expression.
+    # @param ddof [Integer]
+    #   Delta degrees of freedom
+    #
+    # @return [Expr]
+    def pearson_corr(a, b, ddof: 1)
+      if a.is_a?(String)
+        a = col(a)
+      end
+      if b.is_a?(String)
+        b = col(b)
+      end
+      Utils.wrap_expr(RbExpr.pearson_corr(a._rbexpr, b._rbexpr, ddof))
+    end
 
-    # def cov
-    # end
+    # Compute the covariance between two columns/ expressions.
+    #
+    # @param a [Object]
+    #   Column name or Expression.
+    # @param b [Object]
+    #   Column name or Expression.
+    #
+    # @return [Expr]
+    def cov(a, b)
+      if a.is_a?(String)
+        a = col(a)
+      end
+      if b.is_a?(String)
+        b = col(b)
+      end
+      Utils.wrap_expr(RbExpr.cov(a._rbexpr, b._rbexpr))
+    end
 
     # def map
     # end
@@ -408,7 +518,7 @@ module Polars
     #       "bool" => [true, nil],
     #       "list" => [[1, 2], [3]],
     #     }
-    #   ).select([Polars.struct(Polars.all()).alias("my_struct")])
+    #   ).select([Polars.struct(Polars.all).alias("my_struct")])
     #   # =>
     #   # shape: (2, 1)
     #   # ┌─────────────────────┐
@@ -425,7 +535,7 @@ module Polars
     #   df = Polars::DataFrame.new(
     #     {"a" => [1, 2, 3, 4], "b" => ["one", "two", "three", "four"], "c" => [9, 8, 7, 6]}
     #   )
-    #   df.with_column(pl.struct(pl.col(["a", "b"])).alias("a_and_b"))
+    #   df.with_column(Polars.struct(Polars.col(["a", "b"])).alias("a_and_b"))
     #   # =>
     #   # shape: (4, 4)
     #   # ┌─────┬───────┬─────┬─────────────┐

data/lib/polars/lazy_group_by.rb

@@ -1,4 +1,5 @@
 module Polars
+  # Created by `df.lazy.groupby("foo")`.
   class LazyGroupBy
     # @private
     def initialize(lgb, lazyframe_class)