@Test
public void candidatesCountDuplicatesTest() {
for (int parameters : new int[] {1, 2, 5}) {
for (int choices : new int[] {1, 2, 5}) {
for (int candidatesCount : new int[] {10, 100}) {
for (int steps : new int[] {0, 1, 10, 100}) {
try {
// System.out.println("parameters: " + parameters + ", choices: " + choices + ",
// candidatesCount: " + candidatesCount + ", steps: " + steps);
List<List<String>> input = GeneratorTestUtils.prepareInput(parameters, choices);
AdaptiveRandomAlgorithm<String> algorithm =
new AdaptiveRandomAlgorithm<String>(0, candidatesCount, Integer.MAX_VALUE, true);
algorithm.initialize(input, EMPTY_CONSTRAINTS);
for (int i = 0; i < steps; i++) {
algorithm.getNext();
}
List<List<String>> candidates = algorithm.getCandidates();
assertEquals(Math.min(candidatesCount, productSize(input)), candidates.size());
} catch (GeneratorException e) {
fail("Unexpected GeneratorException: " + e.getMessage());
}
}
}
}
}
}
@Test
public void candidatesCountNoDuplicatesTest() {
for (int parameters : new int[] {1, 2, 5}) {
for (int choices : new int[] {1, 2, 5}) {
for (int candidatesCount : new int[] {1, 10, 100}) {
for (int steps : new int[] {0, 1, 10, 100}) {
try {
List<List<String>> input = GeneratorTestUtils.prepareInput(parameters, choices);
AdaptiveRandomAlgorithm<String> algorithm =
new AdaptiveRandomAlgorithm<String>(0, candidatesCount, Integer.MAX_VALUE, false);
algorithm.initialize(input, EMPTY_CONSTRAINTS);
for (int i = 0; i < steps; i++) {
algorithm.getNext();
}
List<List<String>> candidates = algorithm.getCandidates();
assertTrue(candidates.size() <= candidatesCount);
} catch (GeneratorException e) {
fail("Unexpected GeneratorException: " + e.getMessage());
}
}
}
}
}
}
@Test
public void candidatesNoDuplicatesTest() {
for (int parameters : new int[] {1, 2, 5}) {
for (int choices : new int[] {1, 2, 5}) {
for (int candidatesCount : new int[] {1, 10, 100, 1000}) {
try {
List<List<String>> input = GeneratorTestUtils.prepareInput(parameters, choices);
AdaptiveRandomAlgorithm<String> algorithm =
new AdaptiveRandomAlgorithm<String>(0, candidatesCount, Integer.MAX_VALUE, false);
algorithm.initialize(input, EMPTY_CONSTRAINTS);
List<List<String>> candidates = algorithm.getCandidates();
Map<List<String>, Integer> histogram = createHistogram(candidates);
for (int value : histogram.values()) {
assertEquals(1, value);
}
for (List<String> candidate : candidates) {
assertFalse(algorithm.getHistory().contains(candidate));
}
} catch (GeneratorException e) {
fail("Unexpected GeneratorException: " + e.getMessage());
}
}
}
}
}
@Test
public void candidatesDuplicatesTest() {
for (int parameters : new int[] {1, 2, 5}) {
for (int choices : new int[] {1, 2, 5}) {
for (int candidatesCount : new int[] {1, 10, 100, 1000}) {
try {
List<List<String>> input = GeneratorTestUtils.prepareInput(parameters, choices);
AdaptiveRandomAlgorithm<String> algorithm =
new AdaptiveRandomAlgorithm<String>(0, candidatesCount, Integer.MAX_VALUE, true);
algorithm.initialize(input, EMPTY_CONSTRAINTS);
List<List<String>> candidates = algorithm.getCandidates();
assertEquals(candidates.size(), Math.min(candidatesCount, productSize(input)));
} catch (GeneratorException e) {
fail("Unexpected GeneratorException: " + e.getMessage());
}
}
}
}
}
@Test
public void distanceTest() {
AdaptiveRandomAlgorithm<String> algorithm = new AdaptiveRandomAlgorithm<String>(0, 0, 0, false);
List<String> v1 = new ArrayList<String>();
List<String> v2 = new ArrayList<String>();
for (int i = 0; i < 10; ++i) {
String s = GeneratorTestUtils.randomString(10);
v1.add(s);
v2.add(s);
}
int distance = algorithm.distance(v1, v2);
assertEquals(0, distance);
for (int i = 0; i < 10; ++i) {
v1.set(
i,
"some string that is unlikely to occure by random and is also longer than those generated");
assertEquals(i + 1, algorithm.distance(v1, v2));
}
}
@Test
public void candidatesUniformityTest() {
for (int parameters : new int[] {1, 2, 5}) {
for (int choices : new int[] {1, 2, 5}) {
for (int candidatesCount : new int[] {1, 10, 100}) {
for (int steps : new int[] {0, 1, 10, 100, 1000}) {
try {
List<List<String>> input = GeneratorTestUtils.prepareInput(parameters, choices);
AdaptiveRandomAlgorithm<String> algorithm =
new AdaptiveRandomAlgorithm<String>(0, candidatesCount, Integer.MAX_VALUE, true);
algorithm.initialize(input, EMPTY_CONSTRAINTS);
for (int i = 0; i < steps; i++) {
algorithm.getNext();
}
List<List<String>> candidates = algorithm.getCandidates();
if (sampleUniform(candidates, input) == false) {
fail(
"Failed uniformity test for:\n"
+ " parameters = "
+ parameters
+ "\nchoices = "
+ choices
+ "\ncandidatesCount = "
+ candidatesCount
+ "\nsteps = "
+ steps);
}
} catch (GeneratorException e) {
fail("Unexpected GeneratorException: " + e.getMessage());
}
}
}
}
}
}
@Test
public void optimalCandidateTest() {
for (int historySize : new int[] {0, 1, 5, Integer.MAX_VALUE}) {
for (int candidatesSize : new int[] {1, 10, 100}) {
for (int length : new int[] {10000}) {
for (boolean duplicates : new boolean[] {true, false}) {
for (int parameters : new int[] {1, 2, 5}) {
for (int choices : new int[] {1, 2, 5}) {
try {
AdaptiveRandomAlgorithm<String> algorithm =
new AdaptiveRandomAlgorithm<String>(
historySize, candidatesSize, length, duplicates);
algorithm.initialize(
GeneratorTestUtils.prepareInput(parameters, choices), EMPTY_CONSTRAINTS);
List<List<String>> candidates = algorithm.getCandidates();
List<List<String>> history = algorithm.getCandidates();
List<String> optimalCandidate =
algorithm.getOptimalCandidate(candidates, history);
int optimalCandidateDistance = Integer.MAX_VALUE;
for (List<String> event : history) {
optimalCandidateDistance =
Math.min(
algorithm.distance(event, optimalCandidate), optimalCandidateDistance);
}
for (List<String> candidate : candidates) {
int candidateDistance = Integer.MAX_VALUE;
for (List<String> event : history) {
int eventDistance = algorithm.distance(candidate, event);
candidateDistance = Math.min(candidateDistance, eventDistance);
}
assertTrue(candidateDistance <= optimalCandidateDistance);
}
} catch (GeneratorException e) {
fail("Unexpected GeneratorException: " + e.getMessage());
}
}
}
}
}
}
}
}
