/// <summary> /// Use equilateral encoding to encode a column, use zero for the off value and one for the on value. /// <p/> /// http://www.heatonresearch.com/wiki/Equilateral /// </summary> /// <param name="column">The column to use.</param> /// <param name="offValue">The off value to use.</param> /// <param name="onValue">The on value to use.</param> /// <returns>The column to index mapping (the same result as calling enumerateClasses).</returns> public IDictionary <String, int> EncodeEquilateral(int column, double offValue, double onValue) { // remember the column name String name = _headers[column]; // make space for it IDictionary <String, int> classes = EnumerateClasses(column); int classCount = classes.Count; InsertColumns(column + 1, classCount - 1); // perform the equilateral var eq = new Equilateral(classCount, offValue, onValue); foreach (var obj in _data) { int index = classes[obj[column].ToString()]; double[] encoded = eq.Encode(index); for (int i = 0; i < classCount - 1; i++) { obj[column + i] = encoded[i]; } } // name the new columns for (int i = 0; i < classes.Count; i++) { _headers[column + i] = name + "-" + i; } return(classes); }
/// <summary> /// Query a regression algorithm using equilateral encoding. /// </summary> /// <param name="alg">The algorithm being used.</param> /// <param name="theTrainingData">The training data.</param> /// <param name="items">The category items classified.</param> /// <param name="high">The high value.</param> /// <param name="low">The low value.</param> public static void QueryEquilateral( IRegressionAlgorithm alg, IList<BasicData> theTrainingData, IDictionary<String, int> items, double high, double low) { // first, we need to invert the items. Right now it maps from category to index. We need index to category. IDictionary<int, String> invMap = new Dictionary<int, string>(); foreach (string key in items.Keys) { int value = items[key]; invMap[value] = key; } // now we can query Equilateral eq = new Equilateral(items.Count, high, low); foreach (BasicData data in theTrainingData) { double[] output = alg.ComputeRegression(data.Input); int idealIndex = eq.Decode(data.Ideal); int actualIndex = eq.Decode(output); Console.WriteLine(VectorUtil.DoubleArrayToString(data.Input) + " -> " + invMap[actualIndex] + ", Ideal: " + invMap[idealIndex]); } }
/// <summary> /// Use equilateral encoding to encode a column, use zero for the off value and one for the on value. /// <p/> /// http://www.heatonresearch.com/wiki/Equilateral /// </summary> /// <param name="column">The column to use.</param> /// <param name="offValue">The off value to use.</param> /// <param name="onValue">The on value to use.</param> /// <returns>The column to index mapping (the same result as calling enumerateClasses).</returns> public IDictionary<String, int> EncodeEquilateral(int column, double offValue, double onValue) { // remember the column name String name = _headers[column]; // make space for it IDictionary<String, int> classes = EnumerateClasses(column); int classCount = classes.Count; InsertColumns(column + 1, classCount - 1); // perform the equilateral var eq = new Equilateral(classCount, offValue, onValue); foreach (var obj in _data) { int index = classes[obj[column].ToString()]; double[] encoded = eq.Encode(index); for (int i = 0; i < classCount - 1; i++) { obj[column + i] = encoded[i]; } } // name the new columns for (int i = 0; i < classes.Count; i++) { _headers[column + i] = name + "-" + i; } return classes; }