Work in progress.
Able to read short survey, recipe book but crashing in Bayes Net code.
This commit is contained in:
Woody Folsom
218
src/dkohl/bayes/builders/FoodNetBuilder.java
Normal file
218
src/dkohl/bayes/builders/FoodNetBuilder.java
Normal file
@@ -0,0 +1,218 @@
|
||||
package dkohl.bayes.builders;
|
||||
|
||||
import java.util.HashSet;
|
||||
|
||||
import net.woodyfolsom.cs6601.p2.Diner;
|
||||
import net.woodyfolsom.cs6601.p2.Ingredient.TYPE;
|
||||
import net.woodyfolsom.cs6601.p2.Ingredients;
|
||||
import net.woodyfolsom.cs6601.p2.Recipe;
|
||||
import net.woodyfolsom.cs6601.p2.RecipeBook;
|
||||
import net.woodyfolsom.cs6601.p2.Survey;
|
||||
import dkohl.bayes.bayesnet.BayesNet;
|
||||
import dkohl.bayes.estimation.MaximumLikelihoodEstimation;
|
||||
import dkohl.bayes.example.builders.FoodExampleBuilder;
|
||||
import dkohl.bayes.probability.Assignment;
|
||||
import dkohl.bayes.probability.Probability;
|
||||
import dkohl.bayes.probability.Variable;
|
||||
import dkohl.bayes.probability.distribution.ContinousDistribution;
|
||||
import dkohl.bayes.probability.distribution.ProbabilityDistribution;
|
||||
import dkohl.bayes.probability.distribution.ProbabilityTable;
|
||||
import dkohl.bayes.statistic.DataPoint;
|
||||
import dkohl.bayes.statistic.DataSet;
|
||||
import dkohl.onthology.Ontology;
|
||||
|
||||
public class FoodNetBuilder {
|
||||
public static final String TASTE = "Taste";
|
||||
public static final String SOMEONE_VEGETARIAN = "Vegetarian";
|
||||
public static final String CONTAINS_MEAT = "Meat";
|
||||
public static final String CONTAINS_VEGETABLE = "Vegetable";
|
||||
public static final String CONTAINS_BEEF = TYPE.BEEF.toString();
|
||||
public static final String CONTAINS_PORK = TYPE.PORK.toString();
|
||||
public static final String CONTAINS_TOMATOS = TYPE.TOMATO.toString();
|
||||
public static final String CONTAINS_POTATOS = TYPE.POTATO.toString();
|
||||
|
||||
public static final String TRUE_VALUE = "true";
|
||||
public static final String FALSE_VALUE = "false";
|
||||
|
||||
public static final String DOMAIN[] = { TRUE_VALUE, FALSE_VALUE };
|
||||
|
||||
public static final String RATING_DOMAIN[] = { "1", "2", "3", "4", "5",
|
||||
"6", "7", "8", "9", "10" };
|
||||
|
||||
private static final String[] VARIABLES = { SOMEONE_VEGETARIAN,
|
||||
CONTAINS_BEEF, CONTAINS_MEAT, CONTAINS_PORK, CONTAINS_POTATOS,
|
||||
CONTAINS_TOMATOS, CONTAINS_VEGETABLE, TASTE };
|
||||
|
||||
private static final String[] OBSERVED = { CONTAINS_BEEF, CONTAINS_PORK,
|
||||
CONTAINS_POTATOS, CONTAINS_TOMATOS, };
|
||||
|
||||
public static Ontology createOntology() {
|
||||
HashSet<String> classes = new HashSet<String>();
|
||||
|
||||
classes.add(CONTAINS_MEAT);
|
||||
classes.add(CONTAINS_VEGETABLE);
|
||||
|
||||
Ontology onthology = new Ontology(classes);
|
||||
|
||||
onthology.define(CONTAINS_PORK, CONTAINS_MEAT);
|
||||
onthology.define(CONTAINS_BEEF, CONTAINS_MEAT);
|
||||
|
||||
onthology.define(CONTAINS_TOMATOS, CONTAINS_VEGETABLE);
|
||||
onthology.define(CONTAINS_POTATOS, CONTAINS_VEGETABLE);
|
||||
return onthology;
|
||||
}
|
||||
|
||||
private static Assignment build(String varible, String value) {
|
||||
return new Assignment(new Variable(varible, DOMAIN), value);
|
||||
}
|
||||
|
||||
public static DataPoint normalize(DataPoint point, Ontology onto) {
|
||||
// resolve onthology
|
||||
DataPoint normPoint = new DataPoint(point);
|
||||
for (String key : point.keySet()) {
|
||||
if (onto.getInheritance().containsKey(key)) {
|
||||
normPoint
|
||||
.add(build(onto.getInheritance().get(key), TRUE_VALUE));
|
||||
}
|
||||
}
|
||||
|
||||
// implement closed world assumption
|
||||
// everything unknown is false
|
||||
for (String variable : VARIABLES) {
|
||||
if (!normPoint.containsKey(variable)) {
|
||||
normPoint.add(build(variable, FALSE_VALUE));
|
||||
}
|
||||
}
|
||||
return normPoint;
|
||||
}
|
||||
|
||||
public static DataSet getSurveyDataSet(Survey survey, RecipeBook recipeBook) {
|
||||
DataSet data = FoodExampleBuilder.examples();
|
||||
Ontology onto = createOntology();
|
||||
|
||||
int nDishes = survey.getDishCount();
|
||||
for (int dinerIndex = 0; dinerIndex < survey.getDinerCount(); dinerIndex++) {
|
||||
Diner diner = survey.getDiner(dinerIndex);
|
||||
for (int dishIndex = 0; dishIndex < nDishes; dishIndex++) {
|
||||
data.add(normalize(createDataPoint(recipeBook, survey.getDish(dishIndex), diner.getRating(dishIndex)),onto));
|
||||
}
|
||||
}
|
||||
return data;
|
||||
}
|
||||
|
||||
public static DataPoint createDataPoint(RecipeBook recipeBook, String recipeName, int weight) {
|
||||
DataPoint point = new DataPoint();
|
||||
|
||||
Recipe recipe = recipeBook.getRecipe(recipeName);
|
||||
Ingredients ingredients = recipe.getIngredients();
|
||||
|
||||
if (ingredients.contains(TYPE.BEEF)) {
|
||||
point.add(build(CONTAINS_BEEF, TRUE_VALUE));
|
||||
}
|
||||
if (ingredients.contains(TYPE.PORK)) {
|
||||
point.add(build(CONTAINS_PORK, TRUE_VALUE));
|
||||
}
|
||||
if (ingredients.contains(TYPE.POTATO)) {
|
||||
point.add(build(CONTAINS_POTATOS, TRUE_VALUE));
|
||||
}
|
||||
if (ingredients.contains(TYPE.TOMATO)) {
|
||||
point.add(build(CONTAINS_TOMATOS, TRUE_VALUE));
|
||||
}
|
||||
|
||||
point.add(build(TASTE, "" + weight));
|
||||
|
||||
return point;
|
||||
}
|
||||
|
||||
public static ProbabilityDistribution vegi() {
|
||||
String names[] = { SOMEONE_VEGETARIAN };
|
||||
ProbabilityTable table = new ProbabilityTable(names);
|
||||
table.setProbabilityForAssignment("true;", new Probability(0));
|
||||
table.setProbabilityForAssignment("false;", new Probability(1));
|
||||
return table;
|
||||
}
|
||||
|
||||
public static ProbabilityDistribution beef() {
|
||||
String names[] = { CONTAINS_MEAT, CONTAINS_BEEF };
|
||||
ProbabilityTable table = new ProbabilityTable(names);
|
||||
return table;
|
||||
}
|
||||
|
||||
public static ProbabilityDistribution pork() {
|
||||
String names[] = { CONTAINS_MEAT, CONTAINS_PORK };
|
||||
ProbabilityTable table = new ProbabilityTable(names);
|
||||
return table;
|
||||
}
|
||||
|
||||
public static ProbabilityDistribution meet() {
|
||||
String names[] = { SOMEONE_VEGETARIAN, CONTAINS_MEAT };
|
||||
ProbabilityTable table = new ProbabilityTable(names);
|
||||
return table;
|
||||
}
|
||||
|
||||
public static ProbabilityDistribution tomatos() {
|
||||
String names[] = { CONTAINS_VEGETABLE, CONTAINS_TOMATOS };
|
||||
ProbabilityTable table = new ProbabilityTable(names);
|
||||
return table;
|
||||
}
|
||||
|
||||
public static ProbabilityDistribution potatos() {
|
||||
String names[] = { CONTAINS_VEGETABLE, CONTAINS_POTATOS };
|
||||
ProbabilityTable table = new ProbabilityTable(names);
|
||||
return table;
|
||||
}
|
||||
|
||||
public static ProbabilityDistribution vegetables() {
|
||||
String names[] = { CONTAINS_VEGETABLE, SOMEONE_VEGETARIAN };
|
||||
ProbabilityTable table = new ProbabilityTable(names);
|
||||
return table;
|
||||
}
|
||||
|
||||
public static ProbabilityDistribution taste() {
|
||||
String names[] = { TASTE, CONTAINS_BEEF, CONTAINS_PORK,
|
||||
CONTAINS_POTATOS, CONTAINS_TOMATOS };
|
||||
ContinousDistribution distribution = new ContinousDistribution(names, 0);
|
||||
return distribution;
|
||||
}
|
||||
|
||||
public static BayesNet createDishNet(Survey survey, RecipeBook recipeBook) {
|
||||
BayesNet net = new BayesNet(VARIABLES);
|
||||
|
||||
net.setDistribution(new Variable(SOMEONE_VEGETARIAN, DOMAIN), vegi());
|
||||
net.setDistribution(new Variable(CONTAINS_MEAT, DOMAIN), meet());
|
||||
net.setDistribution(new Variable(CONTAINS_VEGETABLE, DOMAIN),
|
||||
vegetables());
|
||||
net.setDistribution(new Variable(CONTAINS_BEEF, DOMAIN), beef());
|
||||
net.setDistribution(new Variable(CONTAINS_PORK, DOMAIN), pork());
|
||||
net.setDistribution(new Variable(CONTAINS_POTATOS, DOMAIN), potatos());
|
||||
net.setDistribution(new Variable(CONTAINS_TOMATOS, DOMAIN), tomatos());
|
||||
net.setDistribution(new Variable(TASTE, RATING_DOMAIN), taste());
|
||||
|
||||
Ontology ontology = createOntology();
|
||||
for (String category : ontology.getClasses()) {
|
||||
net.connect(category, SOMEONE_VEGETARIAN);
|
||||
}
|
||||
|
||||
for (String thing : OBSERVED) {
|
||||
net.connect(thing, ontology.getInheritance().get(thing));
|
||||
net.connect(TASTE, thing);
|
||||
}
|
||||
|
||||
DataSet dataSet = getSurveyDataSet(survey, recipeBook);
|
||||
|
||||
for (String category : ontology.getClasses()) {
|
||||
MaximumLikelihoodEstimation.estimate(dataSet, net, category);
|
||||
for (String thing : ontology.getClasses2thing().get(category)) {
|
||||
MaximumLikelihoodEstimation.estimate(dataSet, net, thing);
|
||||
}
|
||||
}
|
||||
|
||||
MaximumLikelihoodEstimation.estimate(dataSet, net, TASTE);
|
||||
ContinousDistribution distribturion = (ContinousDistribution) net
|
||||
.getNodes().get(TASTE);
|
||||
|
||||
distribturion.estimate();
|
||||
|
||||
return net;
|
||||
}
|
||||
}
|
||||
@@ -27,8 +27,8 @@ public class EnumerateAll {
|
||||
* a set of assignments in this net.
|
||||
* @return
|
||||
*/
|
||||
public static LinkedList<ProbabilityAssignment> enumerateAsk(
|
||||
Variable query, BayesNet net, LinkedList<Assignment> assignments) {
|
||||
public static List<ProbabilityAssignment> enumerateAsk(
|
||||
Variable query, BayesNet net, List<Assignment> assignments) {
|
||||
LinkedList<Variable> variables = net.getVariables();
|
||||
LinkedList<ProbabilityAssignment> result = new LinkedList<ProbabilityAssignment>();
|
||||
|
||||
|
||||
Reference in New Issue
Block a user