From 54e0afae103de0bce86638cee4d8c7d6c498375a Mon Sep 17 00:00:00 2001
From: Wuwei Lin <vincentl13x@gmail.com>
Date: Thu, 28 Jun 2018 20:34:05 +0800
Subject: [PATCH] Add unittest of normone

---
 tests/unit/preprocessor/NormOne_unittest.cc | 65 +++++++++++++++++++++
 1 file changed, 65 insertions(+)
 create mode 100644 tests/unit/preprocessor/NormOne_unittest.cc

diff --git a/tests/unit/preprocessor/NormOne_unittest.cc b/tests/unit/preprocessor/NormOne_unittest.cc
new file mode 100644
index 00000000000..855eb9da64a
--- /dev/null
+++ b/tests/unit/preprocessor/NormOne_unittest.cc
@@ -0,0 +1,65 @@
+/*
+ * This software is distributed under BSD 3-clause license (see LICENSE file).
+ *
+ * Authors: Wuwei Lin
+ */
+
+#include <shogun/mathematics/Math.h>
+#include <shogun/preprocessor/NormOne.h>
+#include <gtest/gtest.h>
+
+using namespace shogun;
+
+class NormOne : public ::testing::Test
+{
+public:
+	NormOne()
+		: feats(Some<CDenseFeatures<float64_t>>::from_raw(nullptr))
+		, transformer(some<CNormOne>())
+	{
+		SGMatrix<float64_t> m(data, num_features, num_vectors, false);
+		m = m.clone();
+		feats = some<CDenseFeatures<float64_t>>(m);
+	}
+
+protected:
+	float64_t data[6] = {1,2,3,4,5,6};
+	float64_t norm [2] = {std::sqrt(1+2*2+3*3), std::sqrt(4*4+5*5+6*6)};
+	int32_t num_vectors = 2;
+	int32_t num_features = 3;
+
+	Some<CDenseFeatures<float64_t>> feats;
+	Some<CNormOne> transformer;
+};
+
+TEST_F(NormOne, transform)
+{
+	transformer->fit(feats);
+	feats = wrap(transformer->transform(feats)->as<CDenseFeatures<float64_t>>());
+
+	EXPECT_EQ(feats->get_num_vectors(), num_vectors);
+
+	for (auto i : range(num_vectors))
+	{
+		SGVector<float64_t> v = feats->get_feature_vector(i);
+		EXPECT_EQ(v.vlen, num_features);
+		for (auto j : range(v.vlen)) {
+			EXPECT_DOUBLE_EQ(v[j], data[num_features*i+j]/norm[i]);
+		}
+	}
+}
+
+TEST_F(NormOne, apply_to_vector)
+{
+	transformer->fit(feats);
+
+	for (auto i : range(num_vectors))
+	{
+		SGVector<float64_t> v = feats->get_feature_vector(i);
+		auto result = transformer->apply_to_feature_vector(v);
+		EXPECT_EQ(result.vlen, num_features);
+		for (auto j : range(v.vlen)) {
+			EXPECT_DOUBLE_EQ(result[j], data[num_features*i+j]/norm[i]);
+		}
+	}
+}