train: @param X: a one dimensional numpy array @param y: a one dimensional numpy array. The contents of y are the labels for the corresponding X values train does not have a return value
(self, X, y)
| 29 | return np.mean((labels - prediction) ** 2) |
| 30 | |
| 31 | def train(self, X, y): |
| 32 | """ |
| 33 | train: |
| 34 | @param X: a one dimensional numpy array |
| 35 | @param y: a one dimensional numpy array. |
| 36 | The contents of y are the labels for the corresponding X values |
| 37 | |
| 38 | train does not have a return value |
| 39 | """ |
| 40 | |
| 41 | """ |
| 42 | this section is to check that the inputs conform to our dimensionality constraints |
| 43 | """ |
| 44 | if X.ndim != 1: |
| 45 | print("Error: Input data set must be one dimensional") |
| 46 | return |
| 47 | if len(X) != len(y): |
| 48 | print("Error: X and y have different lengths") |
| 49 | return |
| 50 | if y.ndim != 1: |
| 51 | print("Error: Data set labels must be one dimensional") |
| 52 | return |
| 53 | |
| 54 | if len(X) < 2 * self.min_leaf_size: |
| 55 | self.prediction = np.mean(y) |
| 56 | return |
| 57 | |
| 58 | if self.depth == 1: |
| 59 | self.prediction = np.mean(y) |
| 60 | return |
| 61 | |
| 62 | best_split = 0 |
| 63 | min_error = self.mean_squared_error(X,np.mean(y)) * 2 |
| 64 | |
| 65 | |
| 66 | """ |
| 67 | loop over all possible splits for the decision tree. find the best split. |
| 68 | if no split exists that is less than 2 * error for the entire array |
| 69 | then the data set is not split and the average for the entire array is used as the predictor |
| 70 | """ |
| 71 | for i in range(len(X)): |
| 72 | if len(X[:i]) < self.min_leaf_size: |
| 73 | continue |
| 74 | elif len(X[i:]) < self.min_leaf_size: |
| 75 | continue |
| 76 | else: |
| 77 | error_left = self.mean_squared_error(X[:i], np.mean(y[:i])) |
| 78 | error_right = self.mean_squared_error(X[i:], np.mean(y[i:])) |
| 79 | error = error_left + error_right |
| 80 | if error < min_error: |
| 81 | best_split = i |
| 82 | min_error = error |
| 83 | |
| 84 | if best_split != 0: |
| 85 | left_X = X[:best_split] |
| 86 | left_y = y[:best_split] |
| 87 | right_X = X[best_split:] |
| 88 | right_y = y[best_split:] |
no test coverage detected