docs: prepare documentation and README for 0.8.0 release

- Bump SPM install example to 0.8.0 in README
- Remove runtime complexity estimates from all doc comments
- Rename 'Combinatorial Extensions' catalog page to 'Mathematical
  Functions' and add missing beta and gamma function entries
- Fix 'Centeral Tendency' filename typo
- Remove empty topic sections from StatKit.md and Distributions.md
- Move FiskDistribution to a dedicated Type Aliases group
- Expand RankingStrategy case doc comments with examples
- Add missing doc comment to NormalDistribution.init(mean:standardDeviation:)
- Add deprecation note to GammaDistribution.init(alpha:beta:)

Author: JimmyMAndersson

README.md

@@ -130,7 +130,7 @@ To install StatKit using the [Swift Package Manager](https://swift.org/package-m
 let package = Package(
     ...
     dependencies: [
-        .package(url: "https://github.com/JimmyMAndersson/StatKit.git", from: "0.6.0")
+        .package(url: "https://github.com/JimmyMAndersson/StatKit.git", from: "0.8.0")
     ],
     ...
 )

Sources/StatKit/Descriptive Statistics/Association/Correlation.swift

@@ -8,7 +8,6 @@ public extension Collection {
   ///
   /// Since there is no notion of correlation in collections with less than two elements,
   /// this method returns NaN if the array count is less than two.
-  /// The time complexity of this method is O(n).
   @inlinable
   func pearsonR<T, U>(
     of X: KeyPath<Element, T>,
@@ -52,7 +51,6 @@ public extension Collection {
   /// Since there is no notion of correlation in collections with less than two
   /// elements or when one of the variables are constant,
   /// this method returns NaN if the array count is less than two.
-  /// The time complexity of this method is O(n).
   @inlinable
   func spearmanR<T, U>(
     of X: KeyPath<Element, T>,
@@ -119,7 +117,6 @@ public extension Collection {
   ///
   /// Since there is no notion of correlation in collections with less than two elements,
   /// this method returns NaN if the array count is less than two.
-  /// The time complexity of this method is O(n).
   func kendallTau<T, U>(
     of X: KeyPath<Element, T>,
     and Y: KeyPath<Element, U>,

Sources/StatKit/Descriptive Statistics/Central Tendency/Averages.swift

@@ -5,7 +5,6 @@ public extension Collection {
   /// - returns: The mean of all items.
   ///
   /// Since the arithmetic mean has no meaning on an empty set, this method returns a NaN if the collection is empty.
-  /// The time complexity of this method is O(n).
   @inlinable
   func mean<T: ConvertibleToReal>(
     variable: KeyPath<Element, T>,
@@ -24,8 +23,6 @@ public extension Collection {
   ///
   /// - important: This method has undefined behavior on collections containing elements that are incomparable.
   /// For example, an array containing NaN's will produce unpredictable results.
-  ///
-  /// - complexity: O(n log n), where n is the length of the collection.
   @inlinable
   func median<T: Comparable & ConvertibleToReal>(
     variable: KeyPath<Element, T>,
@@ -49,8 +46,6 @@ public extension Collection {
   ///
   /// - important: This method has undefined behavior on collections containing elements that are incomparable.
   /// For example, an array containing NaN's will produce unpredictable results.
-  ///
-  /// - complexity: O(n), where n is the length of the collection.
   @inlinable
   func mode<T: Hashable>(
     variable: KeyPath<Element, T>

Sources/StatKit/Descriptive Statistics/Central Tendency/MeanStrategy.swift

@@ -17,7 +17,6 @@ extension MeanStrategy {
   /// - parameter variable: The random variable for which to calculate the mean.
   /// - parameter collection: The collection of values.
   /// - returns: The specified mean value.
-  /// The time complexity of this method is O(n).
   @usableFromInline
   internal func compute<C: Collection, T: ConvertibleToReal>(
     for variable: KeyPath<C.Element, T>,

Sources/StatKit/Descriptive Statistics/Central Tendency/MedianStrategy.swift

@@ -13,8 +13,6 @@ extension MedianStrategy {
   /// - parameter variable: The random variable for which to calculate the mean.
   /// - parameter collection: The collection of values.
   /// - returns: The specified mean value.
-  ///
-  /// - complexity: O(n log n), where n is the length of the collection.
   @usableFromInline
   internal func compute<T, C>(
     for variable: KeyPath<C.Element, T>,

Sources/StatKit/Descriptive Statistics/Dispersion/Range.swift

@@ -6,7 +6,6 @@ public extension Collection {
   /// - returns: The range of the selected variable.
   ///
   /// The range is the difference between the lowest and highest number in a collection.
-  /// The time complexity of this method is O(n).
   func range<T: Comparable & ConvertibleToReal>(
     of variable: KeyPath<Element, T>
   ) -> Double {

Sources/StatKit/Descriptive Statistics/Dispersion/Variability.swift

@@ -5,8 +5,6 @@ public extension Collection {
   /// - parameter variable: The variable over which to calculate the variance.
   /// - parameter composition: The composition of the collection.
   /// - returns: The variance of the collection.
-  ///
-  /// The time complexity of this method is O(n).
   @inlinable
   func variance<T: ConvertibleToReal>(
     variable: KeyPath<Element, T>,
@@ -39,8 +37,6 @@ public extension Collection {
   /// - parameter variable: The variable over which to calculate the standard deviation.
   /// - parameter composition: The composition of the collection.
   /// - returns: The standard deviation of the collection, rounded to a representable value.
-  ///
-  /// The time complexity of this method is O(n).
   @inlinable
   func standardDeviation<T: ConvertibleToReal>(
     variable: KeyPath<Element, T>,
@@ -57,7 +53,6 @@ public extension Collection {
   ///
   /// Since there is no notion of covariance in collections with less than two elements,
   /// this method returns NaN if the array count is less than two.
-  /// The time complexity of this method is O(n).
   @inlinable
   func covariance<T, U>(
     of X: KeyPath<Element, T>,

Sources/StatKit/Descriptive Statistics/Distribution/Continuous/GammaDistribution.swift

@@ -14,6 +14,8 @@ public struct GammaDistribution: ContinuousDistribution, UnivariateDistribution
   /// Creates a Gamma Distribution with a specified alpha and beta parameters.
   /// - parameter alpha: The distribution alpha parameter.
   /// - parameter beta: The distribution beta parameter.
+  ///
+  /// - important: This initializer is deprecated. Use ``init(shape:scale:)`` or ``init(shape:rate:)`` instead.
   @available(*, deprecated, renamed: "init(shape:scale:)")
   public init(alpha: Double, beta: Double) {
     precondition(0 < alpha, "The shape parameter needs to be greater than 0.")

Sources/StatKit/Descriptive Statistics/Distribution/Continuous/NormalDistribution.swift

@@ -18,6 +18,9 @@ public struct NormalDistribution: ContinuousDistribution, UnivariateDistribution
     self.variance = variance
   }
 
+  /// Create a Normal Distribution with a specified mean and standard deviation.
+  /// - parameter mean: The mean value of this distribution.
+  /// - parameter standardDeviation: The standard deviation for this distribution.
   public init(mean: Double, standardDeviation: Double) {
     precondition(
       0 < standardDeviation,

Sources/StatKit/Descriptive Statistics/Ranking/Ranking.swift

@@ -9,8 +9,6 @@ public extension Collection {
   /// - important: This method triggers a precondition failure if attempting to rank a collection
   /// of instances that do not compare equal to themselves. For example, attempting to rank a collection
   /// of floating point numbers that contain NaN's will fail, since `Double.nan == Double.nan` is false.
-  ///
-  /// The time complexity of this method is O(n).
   func rank<T: Comparable & Hashable>(
     variable: KeyPath<Element, T>,
     by order: @escaping (_ lhs: T, _ rhs: T) -> Bool,