[MINOR] Fix python neural network tests (race condition random.seed)

Author: mboehm7

src/main/python/tests/nn/test_neural_network.py

@@ -25,16 +25,13 @@
 from tests.nn.neural_network import NeuralNetwork
 from systemds.script_building.script import DMLScript
 
-# Seed for the input matrix
-np.random.seed(42)
-
-
 class TestNeuralNetwork(unittest.TestCase):
     sds: SystemDSContext = None
 
     @classmethod
     def setUpClass(cls):
         cls.sds = SystemDSContext()
+        np.random.seed(42)
         cls.X = np.random.rand(6, 1)
         cls.exp_out = np.array([
             -0.37768756, -0.47785831, -0.95870362,

src/test/java/org/apache/sysds/test/component/matrix/libMatrixReorg/RollTest.java

@@ -42,82 +42,82 @@
  */
 @RunWith(Parameterized.class)
 public class RollTest {
-    private final int shift;
+	private final int shift;
 
-    // Input matrices
-    private MatrixBlock inputSparse;
-    private MatrixBlock inputDense;
+	// Input matrices
+	private MatrixBlock inputSparse;
+	private MatrixBlock inputDense;
 
-    /**
-     * Constructor for parameterized test cases.
-     *
-     * @param rows     Number of rows in the test matrix.
-     * @param cols     Number of columns in the test matrix.
-     * @param sparsity Sparsity level of the test matrix (0.0 to 1.0).
-     * @param shift    Shift value for the roll operation.
-     */
-    public RollTest(int rows, int cols, double sparsity, int shift) {
-        this.shift = shift;
+	/**
+	 * Constructor for parameterized test cases.
+	 *
+	 * @param rows	 Number of rows in the test matrix.
+	 * @param cols	 Number of columns in the test matrix.
+	 * @param sparsity Sparsity level of the test matrix (0.0 to 1.0).
+	 * @param shift	Shift value for the roll operation.
+	 */
+	public RollTest(int rows, int cols, double sparsity, int shift) {
+		this.shift = shift;
 
-        // Generate a MatrixBlock with the given parameters
-        inputSparse = TestUtils.generateTestMatrixBlock(rows, cols, 0, 10, sparsity, 1);
-        inputSparse.recomputeNonZeros();
+		// Generate a MatrixBlock with the given parameters
+		inputSparse = TestUtils.generateTestMatrixBlock(rows, cols, 0, 10, sparsity, 1);
+		inputSparse.recomputeNonZeros();
 
-        inputDense = new MatrixBlock(rows, cols, false); // false indicates dense
-        inputDense.copy(inputSparse, false); // Copy without maintaining sparsity
-        inputDense.recomputeNonZeros();
-    }
+		inputDense = new MatrixBlock(rows, cols, false); // false indicates dense
+		inputDense.copy(inputSparse, false); // Copy without maintaining sparsity
+		inputDense.recomputeNonZeros();
+	}
 
-    /**
-     * Defines the parameters for the test cases.
-     * Each Object[] contains {rows, cols, sparsity, shift}.
-     *
-     * @return Collection of test parameters.
-     */
-    @Parameters(name = "Rows: {0}, Cols: {1}, Sparsity: {2}, Shift: {3}")
-    public static Collection<Object[]> data() {
-        List<Object[]> tests = new ArrayList<>();
+	/**
+	 * Defines the parameters for the test cases.
+	 * Each Object[] contains {rows, cols, sparsity, shift}.
+	 *
+	 * @return Collection of test parameters.
+	 */
+	@Parameters(name = "Rows: {0}, Cols: {1}, Sparsity: {2}, Shift: {3}")
+	public static Collection<Object[]> data() {
+		List<Object[]> tests = new ArrayList<>();
 
-        // Define various sizes, sparsity levels, and shift values
-        int[] rows = {1, 19, 1001, 2017};
-        int[] cols = {1, 17, 1001, 2017};
-        double[] sparsities = {0.01, 0.1, 0.7, 1.0};
-        int[] shifts = {0, 1, 5, 10, 15};
+		// Define various sizes, sparsity levels, and shift values
+		int[] rows = {1, 19, 1001, 2017};
+		int[] cols = {1, 17, 1001, 2017};
+		double[] sparsities = {0.01, 0.1, 0.7, 1.0};
+		int[] shifts = {0, 1, 5, 10, 15};
 
-        // Generate all combinations of sizes, sparsities, and shifts
-        for (int row : rows) {
-            for (int col : cols) {
-                for (double sparsity : sparsities) {
-                    for (int shift : shifts) {
-                        tests.add(new Object[]{row, col, sparsity, shift});
-                    }
-                }
-            }
-        }
-        return tests;
-    }
+		// Generate all combinations of sizes, sparsities, and shifts
+		for (int row : rows) {
+			for (int col : cols) {
+				for (double sparsity : sparsities) {
+					for (int shift : shifts) {
+						tests.add(new Object[]{row, col, sparsity, shift});
+					}
+				}
+			}
+		}
+		return tests;
+	}
 
-    /**
-     * The actual test method that performs the roll operation on both
-     * sparse and dense matrices and compares the results.
-     */
-    @Test
-    public void test() {
-        try {
-            IndexFunction op = new RollIndex(shift);
-            MatrixBlock outputDense = inputDense.reorgOperations(
+	/**
+	 * The actual test method that performs the roll operation on both
+	 * sparse and dense matrices and compares the results.
+	 */
+	@Test
+	public void test() {
+		try {
+			IndexFunction op = new RollIndex(shift);
+			MatrixBlock outputDense = inputDense.reorgOperations(
 					new ReorgOperator(op), new MatrixBlock(), 0, 0, 0);
-            MatrixBlock outputSparse = inputSparse.reorgOperations(
+			MatrixBlock outputSparse = inputSparse.reorgOperations(
 					new ReorgOperator(op), new MatrixBlock(), 0, 0, 0);
-            outputSparse.sparseToDense();
+			outputSparse.sparseToDense();
 
-            // Compare the dense representations of both outputs
-            TestUtils.compareMatrices(outputSparse, outputDense, 1e-9,
+			// Compare the dense representations of both outputs
+			TestUtils.compareMatrices(outputSparse, outputDense, 1e-9,
 					"Compare Sparse and Dense Roll Results");
 
-        } catch (Exception e) {
-            e.printStackTrace();
-            fail("Exception occurred during roll function test: " + e.getMessage());
-        }
-    }
+		} catch (Exception e) {
+			e.printStackTrace();
+			fail("Exception occurred during roll function test: " + e.getMessage());
+		}
+	}
 }