attribute.java

wekaUT是 university texas austin 开发的基于weka的半指导学习(semi supervised learning)的分类器

   * Tests if the attribute is a string.
   *
   * @return true if the attribute is a string
   */
  public final boolean isString() {

    return (m_Type == STRING);
  }

  /**
   * Tests if the attribute is a date type.
   *
   * @return true if the attribute is a date type
   */
  public final boolean isDate() {

    return (m_Type == DATE);
  }

  /**
   * Returns the attribute's name.
   *
   * @return the attribute's name as a string
   */
  public final String name() {

    return m_Name;
  }
  
  /**
   * Returns the number of attribute values. Returns 0 for numeric attributes.
   *
   * @return the number of attribute values
   */
  public final int numValues() {

    if (!isNominal() && !isString()) {
      return 0;
    } else {
      return m_Values.size();
    }
  }

  /**
   * Returns a description of this attribute in ARFF format. Quotes
   * strings if they contain whitespace characters, or if they
   * are a question mark.
   *
   * @return a description of this attribute as a string
   */
  public final String toString() {
    
    StringBuffer text = new StringBuffer();
    
    text.append("@attribute " + Utils.quote(m_Name) + " ");
    switch (m_Type) {
    case NOMINAL:
      text.append('{');
      Enumeration enum = enumerateValues();
      while (enum.hasMoreElements()) {
	text.append(Utils.quote((String) enum.nextElement()));
	if (enum.hasMoreElements())
	  text.append(',');
      }
      text.append('}');
      break;
    case NUMERIC:
      text.append("numeric");
      break;
    case STRING:
      text.append("string");
      break;
    case DATE:
      text.append("date ").append(Utils.quote(m_DateFormat.toPattern()));
      break;
    default:
      text.append("UNKNOWN");
      break;
    }
    return text.toString();
  }

  /**
   * Returns the attribute's type as an integer.
   *
   * @return the attribute's type.
   */
  public final int type() {

    return m_Type;
  }

  /**
   * Returns a value of a nominal or string attribute. 
   * Returns an empty string if the attribute is neither
   * nominal nor a string attribute.
   *
   * @param valIndex the value's index
   * @return the attribute's value as a string
   */
  public final String value(int valIndex) {
    
    if (!isNominal() && !isString()) {
      return "";
    } else {
      Object val = m_Values.elementAt(valIndex);
      
      // If we're storing strings compressed, uncompress it.
      if (val instanceof SerializedObject) {
        val = ((SerializedObject)val).getObject();
      }
      return (String) val;
    }
  }

  /**
   * Constructor for a numeric attribute with a particular index.
   *
   * @param attributeName the name for the attribute
   * @param index the attribute's index
   */
  Attribute(String attributeName, int index) {

    this(attributeName);
    m_Index = index;
  }

  /**
   * Constructor for date attributes with a particular index.
   *
   * @param attributeName the name for the attribute
   * @param dateFormat a string suitable for use with
   * SimpleDateFormatter for parsing dates.  Null for a default format
   * string.
   * @param index the attribute's index
   */
  Attribute(String attributeName, String dateFormat, 
	    int index) {

    this(attributeName, dateFormat);
    m_Index = index;
  }

  /**
   * Constructor for nominal attributes and string attributes with
   * a particular index.
   * If a null vector of attribute values is passed to the method,
   * the attribute is assumed to be a string.
   *
   * @param attributeName the name for the attribute
   * @param attributeValues a vector of strings denoting the attribute values.
   * Null if the attribute is a string attribute.
   * @param index the attribute's index
   */
  Attribute(String attributeName, FastVector attributeValues, 
	    int index) {

    this(attributeName, attributeValues);
    m_Index = index;
  }

  /**
   * Adds a string value to the list of valid strings for attributes
   * of type STRING and returns the index of the string.
   *
   * @param value The string value to add
   * @return the index assigned to the string, or -1 if the attribute is not
   * of type Attribute.STRING 
   */
  public int addStringValue(String value) {

    if (!isString()) {
      return -1;
    }
    Object store = value;

    if (value.length() > STRING_COMPRESS_THRESHOLD) {
      try {
        store = new SerializedObject(value, true);
      } catch (Exception ex) {
        System.err.println("Couldn't compress string attribute value -"
                           + " storing uncompressed.");
      }
    }
    Integer index = (Integer)m_Hashtable.get(store);
    if (index != null) {
      return index.intValue();
    } else {
      int intIndex = m_Values.size();
      m_Values.addElement(store);
      m_Hashtable.put(store, new Integer(intIndex));
      return intIndex;
    }
  }

  /**
   * Adds a string value to the list of valid strings for attributes
   * of type STRING and returns the index of the string. This method is
   * more efficient than addStringValue(String) for long strings.
   *
   * @param src The Attribute containing the string value to add.
   * @param int index the index of the string value in the source attribute.
   * @return the index assigned to the string, or -1 if the attribute is not
   * of type Attribute.STRING 
   */
  public int addStringValue(Attribute src, int index) {

    if (!isString()) {
      return -1;
    }
    Object store = src.m_Values.elementAt(index);
    Integer oldIndex = (Integer)m_Hashtable.get(store);
    if (oldIndex != null) {
      return oldIndex.intValue();
    } else {
      int intIndex = m_Values.size();
      m_Values.addElement(store);
      m_Hashtable.put(store, new Integer(intIndex));
      return intIndex;
    }
  }

  /**
   * Adds an attribute value. Creates a fresh list of attribute
   * values before adding it.
   *
   * @param value the attribute value
   */
  final void addValue(String value) {

    m_Values = (FastVector)m_Values.copy();
    m_Hashtable = (Hashtable)m_Hashtable.clone();
    forceAddValue(value);
  }

  /**
   * Produces a shallow copy of this attribute with a new name.
   *
   * @param newName the name of the new attribute
   * @return a copy of this attribute with the same index
   */
  final Attribute copy(String newName) {

    Attribute copy = new Attribute(newName);

    copy.m_Index = m_Index;
    copy.m_DateFormat = m_DateFormat;
    copy.m_Type = m_Type;
    copy.m_Values = m_Values;
    copy.m_Hashtable = m_Hashtable;
    copy.setMetadata(m_Metadata);
 
    return copy;
  }

  /**
   * Removes a value of a nominal or string attribute. Creates a 
   * fresh list of attribute values before removing it.
   *
   * @param index the value's index
   * @exception IllegalArgumentException if the attribute is not nominal
   */
  final void delete(int index) {
    
    if (!isNominal() && !isString()) 
      throw new IllegalArgumentException("Can only remove value of" +
                                         "nominal or string attribute!");
    else {
      m_Values = (FastVector)m_Values.copy();
      m_Values.removeElementAt(index);
      Hashtable hash = new Hashtable(m_Hashtable.size());
      Enumeration enum = m_Hashtable.keys();
      while (enum.hasMoreElements()) {
	Object string = enum.nextElement();
	Integer valIndexObject = (Integer)m_Hashtable.get(string);
	int valIndex = valIndexObject.intValue();
	if (valIndex > index) {
	  hash.put(string, new Integer(valIndex - 1));
	} else if (valIndex < index) {
	  hash.put(string, valIndexObject);
	}
      }
      m_Hashtable = hash;
    }
  }

  /**
   * Adds an attribute value.
   *
   * @param value the attribute value
   */
  final void forceAddValue(String value) {

    Object store = value;
    if (value.length() > STRING_COMPRESS_THRESHOLD) {
      try {
        store = new SerializedObject(value, true);
      } catch (Exception ex) {
        System.err.println("Couldn't compress string attribute value -"
                           + " storing uncompressed.");
      }
    }
    m_Values.addElement(store);
    m_Hashtable.put(store, new Integer(m_Values.size() - 1));
  }

  /**
   * Sets the index of this attribute.
   *
   * @param the index of this attribute
   */
  final void setIndex(int index) {

    m_Index = index;
  }

  /**
   * Sets a value of a nominal attribute or string attribute.
   * Creates a fresh list of attribute values before it is set.
   *
   * @param index the value's index
   * @param string the value
   * @exception IllegalArgumentException if the attribute is not nominal or 
   * string.
   */
  final void setValue(int index, String string) {
    
    switch (m_Type) {
    case NOMINAL:
    case STRING:
      m_Values = (FastVector)m_Values.copy();
      m_Hashtable = (Hashtable)m_Hashtable.clone();
      Object store = string;
      if (string.length() > STRING_COMPRESS_THRESHOLD) {
        try {
          store = new SerializedObject(string, true);
        } catch (Exception ex) {
          System.err.println("Couldn't compress string attribute value -"
                             + " storing uncompressed.");
        }
      }
      m_Hashtable.remove(m_Values.elementAt(index));
      m_Values.setElementAt(store, index);
      m_Hashtable.put(store, new Integer(index));
      break;
    default:
      throw new IllegalArgumentException("Can only set values for nominal"
                                         + " or string attributes!");
    }
  }

  public String formatDate(double date) {
    switch (m_Type) {
    case DATE:
      return m_DateFormat.format(new Date((long)date));
    default:
      throw new IllegalArgumentException("Can only format date values for date"
                                         + " attributes!");
    }
  }

  public double parseDate(String string) throws ParseException {
    switch (m_Type) {
    case DATE:
      long time = m_DateFormat.parse(string).getTime();
      // TODO put in a safety check here if we can't store the value in a double.
      return (double)time;
    default:
      throw new IllegalArgumentException("Can only parse date values for date"
                                         + " attributes!");
    }
  }

  /**
   * Returns the properties supplied for this attribute.
   *
   * @return metadata for this attribute
   */  
  public final ProtectedProperties getMetadata() {

    return m_Metadata;
  }

  /**
   * Returns the ordering of the attribute. One of the following:
   * 
   * ORDERING_SYMBOLIC - attribute values should be treated as symbols.
   * ORDERING_ORDERED  - attribute values have a global ordering.
   * ORDERING_MODULO   - attribute values have an ordering which wraps.
   *
   * @return the ordering type of the attribute
   */
  public final int ordering() {

    return m_Ordering;
  }

  /**
   * Returns whether the attribute values are equally spaced.
   *
   * @return whether the attribute is regular or not
   */
  public final boolean isRegular() {

    return m_IsRegular;
  }