diff --git a/src/FSharp/Differentiate.fs b/src/FSharp/Differentiate.fs
new file mode 100644
index 00000000..1880e6ba
--- /dev/null
+++ b/src/FSharp/Differentiate.fs
@@ -0,0 +1,63 @@
+// <copyright file="Differentiate.fs" company="Math.NET">
+// Math.NET Numerics, part of the Math.NET Project
+// http://numerics.mathdotnet.com
+// http://github.com/mathnet/mathnet-numerics
+// http://mathnetnumerics.codeplex.com
+//
+// Copyright (c) 2009-2015 Math.NET
+//
+// Permission is hereby granted, free of charge, to any person
+// obtaining a copy of this software and associated documentation
+// files (the "Software"), to deal in the Software without
+// restriction, including without limitation the rights to use,
+// copy, modify, merge, publish, distribute, sublicense, and/or sell
+// copies of the Software, and to permit persons to whom the
+// Software is furnished to do so, subject to the following
+// conditions:
+//
+// The above copyright notice and this permission notice shall be
+// included in all copies or substantial portions of the Software.
+//
+// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+// EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES
+// OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+// NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
+// HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
+// WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+// FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+// OTHER DEALINGS IN THE SOFTWARE.
+// </copyright>
+
+namespace MathNet.Numerics
+
+open System
+open MathNet.Numerics
+open MathNet.Numerics.Differentiation
+
+[<CompilationRepresentation(CompilationRepresentationFlags.ModuleSuffix)>]
+module Differentiate =
+
+    let private tobcl (f:'a->'b) = Func<'a,'b>(f)
+    let private tofs (f:Func<_,_>) = fun a -> f.Invoke(a)
+    let private tofs2 (f:Func<_,_>) = fun a b -> f.Invoke([|a;b|])
+
+    let derivative order x (f:float->float) = Differentiate.Derivative(tobcl f, x, order)
+    let derivativeFunc order (f:float->float) = Differentiate.DerivativeFunc(tobcl f, order) |> tofs
+
+    let firstDerivative x (f:float->float) = Differentiate.FirstDerivative(tobcl f, x)
+    let firstDerivativeFunc (f:float->float) = Differentiate.FirstDerivativeFunc(tobcl f) |> tofs
+
+    let secondDerivative x (f:float->float) = Differentiate.SecondDerivative(tobcl f, x)
+    let secondDerivativeFunc (f:float->float) = Differentiate.SecondDerivativeFunc(tobcl f) |> tofs
+
+    let partialDerivative order parameterIndex x (f:float[]->float) = Differentiate.PartialDerivative(tobcl f, x, parameterIndex, order)
+    let partialDerivativeFunc order parameterIndex (f:float[]->float) = Differentiate.PartialDerivativeFunc(tobcl f, parameterIndex, order) |> tofs
+
+    let firstPartialDerivative parameterIndex x (f:float[]->float) = Differentiate.FirstPartialDerivative(tobcl f, x, parameterIndex)
+    let firstPartialDerivativeFunc parameterIndex (f:float[]->float) = Differentiate.FirstPartialDerivativeFunc(tobcl f, parameterIndex) |> tofs
+
+    let partialDerivative2 order parameterIndex x (f:float->float->float) = Differentiate.PartialDerivative((fun x -> f x.[0] x.[1]), x, parameterIndex, order)
+    let partialDerivative2Func order parameterIndex (f:float->float->float) = Differentiate.PartialDerivativeFunc((fun x -> f x.[0] x.[1]), parameterIndex, order) |> tofs2
+
+    let firstPartialDerivative2 parameterIndex x (f:float->float->float) = Differentiate.FirstPartialDerivative((fun x -> f x.[0] x.[1]), x, parameterIndex)
+    let firstPartialDerivative2Func parameterIndex (f:float->float->float) = Differentiate.FirstPartialDerivativeFunc((fun x -> f x.[0] x.[1]), parameterIndex) |> tofs2
diff --git a/src/FSharp/FSharp-Net35.fsproj b/src/FSharp/FSharp-Net35.fsproj
index b3284f9b..9c60efb9 100644
--- a/src/FSharp/FSharp-Net35.fsproj
+++ b/src/FSharp/FSharp-Net35.fsproj
@@ -70,6 +70,7 @@
     <Compile Include="BigIntegerExtensions.fs" />
     <Compile Include="BigRational.fsi" />
     <Compile Include="BigRational.fs" />
+    <Compile Include="Differentiate.fs" />
     <Compile Include="Fit.fs" />
     <Compile Include="FindRoots.fs" />
     <Compile Include="RandomVariable.fs" />
diff --git a/src/FSharp/FSharp-Portable47.fsproj b/src/FSharp/FSharp-Portable47.fsproj
index 16fe7d25..dbd09e80 100644
--- a/src/FSharp/FSharp-Portable47.fsproj
+++ b/src/FSharp/FSharp-Portable47.fsproj
@@ -58,6 +58,7 @@
     <Compile Include="BigIntegerExtensions.fs" />
     <Compile Include="BigRational.fsi" />
     <Compile Include="BigRational.fs" />
+    <Compile Include="Differentiate.fs" />
     <Compile Include="Fit.fs" />
     <Compile Include="FindRoots.fs" />
     <Compile Include="RandomVariable.fs" />
diff --git a/src/FSharp/FSharp.fsproj b/src/FSharp/FSharp.fsproj
index e40ed8c8..3023643f 100644
--- a/src/FSharp/FSharp.fsproj
+++ b/src/FSharp/FSharp.fsproj
@@ -70,6 +70,7 @@
     <Compile Include="BigIntegerExtensions.fs" />
     <Compile Include="BigRational.fsi" />
     <Compile Include="BigRational.fs" />
+    <Compile Include="Differentiate.fs" />
     <Compile Include="Fit.fs" />
     <Compile Include="FindRoots.fs" />
     <Compile Include="RandomVariable.fs" />
diff --git a/src/FSharp/FindRoots.fs b/src/FSharp/FindRoots.fs
index 30e4c30e..53c8a730 100644
--- a/src/FSharp/FindRoots.fs
+++ b/src/FSharp/FindRoots.fs
@@ -37,46 +37,46 @@ open MathNet.Numerics.RootFinding
 module FindRoots =
 
     let private tobcl (f:'a->'b) = Func<'a,'b>(f)
-    let private (|>|) option orElse = match option with | Some x -> Some x | None -> orElse() 
+    let private (|>|) option orElse = match option with | Some x -> Some x | None -> orElse()
 
     // direct algorithms
 
-    let bisection maxIterations accuracy lowerBound upperBound (f:double->double) =
+    let bisection maxIterations accuracy lowerBound upperBound (f:float->float) =
         match Bisection.TryFindRoot(tobcl f, lowerBound, upperBound, accuracy, maxIterations) with
         | true, root -> Some root
         | false, _ -> None
 
-    let brent maxIterations accuracy lowerBound upperBound (f:double->double) =
+    let brent maxIterations accuracy lowerBound upperBound (f:float->float) =
         match Brent.TryFindRoot(tobcl f, lowerBound, upperBound, accuracy, maxIterations) with
         | true, root -> Some root
         | false, _ -> None
 
-    let newtonRaphson maxIterations accuracy lowerBound upperBound (f:double->double) (df:double->double) =
+    let newtonRaphson maxIterations accuracy lowerBound upperBound (f:float->float) (df:float->float) =
         match NewtonRaphson.TryFindRoot(tobcl f, tobcl df, 0.5 * (lowerBound + upperBound), lowerBound, upperBound, accuracy, maxIterations) with
         | true, root -> Some root
         | false, _ -> None
 
-    let newtonRaphsonGuess maxIterations accuracy guess (f:double->double) (df:double->double) =
+    let newtonRaphsonGuess maxIterations accuracy guess (f:float->float) (df:float->float) =
         match NewtonRaphson.TryFindRoot(tobcl f, tobcl df, guess, Double.MinValue, Double.MaxValue, accuracy, maxIterations) with
         | true, root -> Some root
         | false, _ -> None
 
-    let newtonRaphsonRobust maxIterations subdivision accuracy lowerBound upperBound (f:double->double) (df:double->double) =
+    let newtonRaphsonRobust maxIterations subdivision accuracy lowerBound upperBound (f:float->float) (df:float->float) =
         match RobustNewtonRaphson.TryFindRoot(tobcl f, tobcl df, lowerBound, upperBound, accuracy, maxIterations, subdivision) with
         | true, root -> Some root
         | false, _ -> None
 
-    let broyden maxIterations accuracy guess (f:double[]->double[]) =
+    let broyden maxIterations accuracy guess (f:float[]->float[]) =
         match Broyden.TryFindRoot(tobcl f, guess, accuracy, maxIterations) with
         | true, root -> Some root
         | false, _ -> None
 
     // simple usage
 
-    let ofFunction lowerBound upperBound (f:double->double) =
+    let ofFunction lowerBound upperBound (f:float->float) =
         brent 100 1e-8 lowerBound upperBound f
         |>| fun () -> bisection 100 1e-8 lowerBound upperBound f
 
-    let ofFunctionDerivative lowerBound upperBound (f:double->double) (df:double->double) =
+    let ofFunctionDerivative lowerBound upperBound (f:float->float) (df:float->float) =
         newtonRaphsonRobust 100 20 1e-8 lowerBound upperBound f df
         |>| fun () -> bisection 100 1e-8 lowerBound upperBound f
diff --git a/src/FSharp/Random.fs b/src/FSharp/Random.fs
index 7731f501..342f03fc 100644
--- a/src/FSharp/Random.fs
+++ b/src/FSharp/Random.fs
@@ -39,11 +39,11 @@ module Random =
     /// Default sampling, efficient but without custom seed (uses robust seeds internally)
     let inline doubles (length:int) = SystemRandomSource.FastDoubles(length)
     let inline doubleSeq () = SystemRandomSource.DoubleSequence()
-    let inline doubleFill (values:double[]) = SystemRandomSource.FastDoubles(values)
+    let inline doubleFill (values:float[]) = SystemRandomSource.FastDoubles(values)
 
     let inline doublesSeed (seed:int) (length:int) = SystemRandomSource.Doubles(length, seed)
     let inline doubleSeqSeed (seed:int) = SystemRandomSource.DoubleSequence(seed)
-    let inline doubleFillSeed (seed:int) (values:double[]) = SystemRandomSource.Doubles(values, seed)
+    let inline doubleFillSeed (seed:int) (values:float[]) = SystemRandomSource.Doubles(values, seed)
 
     /// Creates a default .Net system pRNG with a robust seed
     let systemShared = shared