Prevent DataFrame.Sample() method from returning duplicated rows (#2939)

* resolves #2806

* replace forloop with ArraySegment<T>

* reduce shuffle loop operations from O(Rows.Count) to O(numberOfRows)

Author: RamonWill

src/Microsoft.Data.Analysis/DataFrame.cs

@@ -328,14 +328,28 @@ public DataFrame AddSuffix(string suffix, bool inPlace = false)
         /// <param name="numberOfRows">Number of rows in the returned DataFrame</param>
         public DataFrame Sample(int numberOfRows)
         {
+            if (numberOfRows > Rows.Count)
+            {
+                throw new ArgumentException(string.Format(Strings.ExceedsNumberOfRows, Rows.Count), nameof(numberOfRows));
+            }
+
+            int shuffleLowerLimit = 0;
+            int shuffleUpperLimit = (int)Math.Min(Int32.MaxValue, Rows.Count);
+            
+            int[] shuffleArray = Enumerable.Range(0, shuffleUpperLimit).ToArray();
             Random rand = new Random();
-            PrimitiveDataFrameColumn<long> indices = new PrimitiveDataFrameColumn<long>("Indices", numberOfRows);
-            int randMaxValue = (int)Math.Min(Int32.MaxValue, Rows.Count);
-            for (long i = 0; i < numberOfRows; i++)
+            while (shuffleLowerLimit < numberOfRows)
             {
-                indices[i] = rand.Next(randMaxValue);
+                int randomIndex = rand.Next(shuffleLowerLimit, shuffleUpperLimit);
+                int temp = shuffleArray[shuffleLowerLimit];
+                shuffleArray[shuffleLowerLimit] = shuffleArray[randomIndex];
+                shuffleArray[randomIndex] = temp;
+                shuffleLowerLimit++;
             }
+            ArraySegment<int> segment = new ArraySegment<int>(shuffleArray, 0, shuffleLowerLimit);
 
+            PrimitiveDataFrameColumn<int> indices = new PrimitiveDataFrameColumn<int>("indices", segment);
+            
             return Clone(indices);
         }
 

src/Microsoft.Data.Analysis/strings.Designer.cs

@@ -141,6 +141,9 @@
   <data name="ExceedsNumberOfColumns" xml:space="preserve">
     <value>Parameter.Count exceeds the number of columns({0}) in the DataFrame </value>
   </data>
+  <data name="ExceedsNumberOfRows" xml:space="preserve">
+    <value>Parameter.Count exceeds the number of rows({0}) in the DataFrame </value>
+  </data>
   <data name="ExpectedEitherGuessRowsOrDataTypes" xml:space="preserve">
     <value>Expected either {0} or {1} to be provided</value>
   </data>
@@ -186,4 +189,4 @@
   <data name="SpansMultipleBuffers" xml:space="preserve">
     <value>Cannot span multiple buffers</value>
   </data>
-</root>
+</root>

src/Microsoft.Data.Analysis/strings.resx

@@ -1561,9 +1561,20 @@ public void TestPrefixAndSuffix()
         public void TestSample()
         {
             DataFrame df = MakeDataFrameWithAllColumnTypes(10);
-            DataFrame sampled = df.Sample(3);
-            Assert.Equal(3, sampled.Rows.Count);
+            DataFrame sampled = df.Sample(7);
+            Assert.Equal(7, sampled.Rows.Count);
             Assert.Equal(df.Columns.Count, sampled.Columns.Count);
+
+            // all sampled rows should be unique.
+            HashSet<int?> uniqueRowValues = new HashSet<int?>();
+            foreach(int? value in sampled.Columns["Int"])
+            {
+                uniqueRowValues.Add(value);
+            }
+            Assert.Equal(uniqueRowValues.Count, sampled.Rows.Count);
+
+            // should throw exception as sample size is greater than dataframe rows
+            Assert.Throws<ArgumentException>(()=> df.Sample(13));
         }
 
         [Fact]