fix: resolve stack overflow in noncentral_f and Int64 test mismatches

- Fix circular delegation in np.random.noncentral_f: params long[] now has
  the implementation instead of delegating to Shape (which delegated back)
- Fix np.random.f params long[] to delegate to int[] instead of Shape
- Update np.split tests to use ToArray<long>() for np.arange results
  (np.arange now returns Int64 in longindexing branch)
- Fix Split_ReturnsViews test to use Get/SetInt64 instead of Int32

Author: Nucs

src/NumSharp.Core/RandomSampling/np.random.f.cs

@@ -92,7 +92,7 @@ public NDArray f(double dfnum, double dfden, int[] size)
         /// <returns>Drawn samples from the F distribution.</returns>
         /// <exception cref="ArgumentException">If dfnum or dfden is <= 0.</exception>
         public NDArray f(double dfnum, double dfden, params long[] size)
-            => f(dfnum, dfden, new Shape(size));
+            => f(dfnum, dfden, Array.ConvertAll(size, d => (int)d));
 
         /// <summary>
         ///     Generate a single sample from the F distribution.

src/NumSharp.Core/RandomSampling/np.random.noncentral_f.cs

@@ -37,6 +37,20 @@ public NDArray noncentral_f(double dfnum, double dfden, double nonc, Shape size)
         ///     https://numpy.org/doc/stable/reference/random/generated/numpy.random.noncentral_f.html
         /// </remarks>
         public NDArray noncentral_f(double dfnum, double dfden, double nonc, int[] size)
+            => noncentral_f(dfnum, dfden, nonc, Shape.ComputeLongShape(size));
+
+        /// <summary>
+        ///     Draw samples from the noncentral F distribution.
+        /// </summary>
+        /// <param name="dfnum">Numerator degrees of freedom, must be > 0.</param>
+        /// <param name="dfden">Denominator degrees of freedom, must be > 0.</param>
+        /// <param name="nonc">Non-centrality parameter, must be >= 0.</param>
+        /// <param name="size">Output shape.</param>
+        /// <returns>Drawn samples from the parameterized noncentral F distribution.</returns>
+        /// <remarks>
+        ///     https://numpy.org/doc/stable/reference/random/generated/numpy.random.noncentral_f.html
+        /// </remarks>
+        public NDArray noncentral_f(double dfnum, double dfden, double nonc, params long[] size)
         {
             // Parameter validation (matches NumPy error messages)
             if (dfnum <= 0)
@@ -51,30 +65,19 @@ public NDArray noncentral_f(double dfnum, double dfden, double nonc, int[] size)
                 return NDArray.Scalar(SampleNoncentralF(dfnum, dfden, nonc));
             }
 
-            var result = new NDArray<double>(size);
-            ArraySlice<double> resultArray = result.Data<double>();
+            var shape = new Shape(size);
+            var result = new NDArray(NPTypeCode.Double, shape, false);
 
-            for (int i = 0; i < result.size; ++i)
-                resultArray[i] = SampleNoncentralF(dfnum, dfden, nonc);
+            unsafe
+            {
+                var addr = (double*)result.Address;
+                for (long i = 0; i < result.size; ++i)
+                    addr[i] = SampleNoncentralF(dfnum, dfden, nonc);
+            }
 
-            result.ReplaceData(resultArray);
             return result;
         }
 
-        /// <summary>
-        ///     Draw samples from the noncentral F distribution.
-        /// </summary>
-        /// <param name="dfnum">Numerator degrees of freedom, must be > 0.</param>
-        /// <param name="dfden">Denominator degrees of freedom, must be > 0.</param>
-        /// <param name="nonc">Non-centrality parameter, must be >= 0.</param>
-        /// <param name="size">Output shape.</param>
-        /// <returns>Drawn samples from the parameterized noncentral F distribution.</returns>
-        /// <remarks>
-        ///     https://numpy.org/doc/stable/reference/random/generated/numpy.random.noncentral_f.html
-        /// </remarks>
-        public NDArray noncentral_f(double dfnum, double dfden, double nonc, params long[] size)
-            => noncentral_f(dfnum, dfden, nonc, new Shape(size));
-
         /// <summary>
         ///     Sample a single value from the noncentral F distribution.
         /// </summary>

test/NumSharp.UnitTest/Manipulation/np.split.Tests.cs

@@ -21,9 +21,9 @@ public void Split_1D_EqualParts()
             var result = np.split(a, 3);
 
             result.Length.Should().Be(3);
-            result[0].ToArray<int>().Should().ContainInOrder(0, 1, 2);
-            result[1].ToArray<int>().Should().ContainInOrder(3, 4, 5);
-            result[2].ToArray<int>().Should().ContainInOrder(6, 7, 8);
+            result[0].ToArray<long>().Should().ContainInOrder(0, 1, 2);
+            result[1].ToArray<long>().Should().ContainInOrder(3, 4, 5);
+            result[2].ToArray<long>().Should().ContainInOrder(6, 7, 8);
         }
 
         [Test]
@@ -61,12 +61,12 @@ public void Split_2D_Axis1()
             result[1].Should().BeShaped(3, 2);
 
             // First result: columns 0-1
-            result[0]["0, :"].ToArray<int>().Should().ContainInOrder(0, 1);
-            result[0]["1, :"].ToArray<int>().Should().ContainInOrder(4, 5);
+            result[0]["0, :"].ToArray<long>().Should().ContainInOrder(0, 1);
+            result[0]["1, :"].ToArray<long>().Should().ContainInOrder(4, 5);
 
             // Second result: columns 2-3
-            result[1]["0, :"].ToArray<int>().Should().ContainInOrder(2, 3);
-            result[1]["1, :"].ToArray<int>().Should().ContainInOrder(6, 7);
+            result[1]["0, :"].ToArray<long>().Should().ContainInOrder(2, 3);
+            result[1]["1, :"].ToArray<long>().Should().ContainInOrder(6, 7);
         }
 
         [Test]
@@ -96,10 +96,10 @@ public void Split_WithIndices()
             var result = np.split(a, new[] { 3, 5, 7 });
 
             result.Length.Should().Be(4);
-            result[0].ToArray<int>().Should().ContainInOrder(0, 1, 2);
-            result[1].ToArray<int>().Should().ContainInOrder(3, 4);
-            result[2].ToArray<int>().Should().ContainInOrder(5, 6);
-            result[3].ToArray<int>().Should().ContainInOrder(7, 8);
+            result[0].ToArray<long>().Should().ContainInOrder(0, 1, 2);
+            result[1].ToArray<long>().Should().ContainInOrder(3, 4);
+            result[2].ToArray<long>().Should().ContainInOrder(5, 6);
+            result[3].ToArray<long>().Should().ContainInOrder(7, 8);
         }
 
         [Test]
@@ -295,8 +295,8 @@ public void Split_ReturnsViews()
             var result = np.split(a, 3);
 
             // Modifying a view should affect the original
-            result[0].SetInt32(999, 0);
-            a.GetInt32(0).Should().Be(999);
+            result[0].SetInt64(999, 0);
+            a.GetInt64(0).Should().Be(999);
         }
 
         [Test]