Merge branch 'master' of github.com:rpmcruz/machine-learning

timeseries/delay.py

@@ -0,0 +1,10 @@
+# introduce delay in the target variable
+
+import numpy as np
+
+def shift(y, delay):
+    X = np.asarray([np.roll(y, d) for d in range(1, delay+1)]).T
+    X = X[delay:]
+    y = y[delay:]
+    t = np.arange(delay, len(y)+delay, dtype=int)
+    return t, X, y

timeseries/growing.py

@@ -0,0 +1,12 @@
+import numpy as np
+
+class GrowingWindow:
+    def __init__(self, n_splits=3, test_size=1):
+        self.n_splits = n_splits
+        self.test_size = test_size
+
+    def split(self, X, y=None):
+        train_size = len(X)-self.test_size
+        train_blocks = np.linspace(0, train_size, self.n_splits+1, dtype=int)
+        for j in train_blocks[1:]:
+            yield np.arange(0, j), np.arange(j, j+self.test_size)

timeseries/sliding.py

@@ -0,0 +1,14 @@
+import numpy as np
+
+class SlidingWindow:
+    def __init__(self, n_splits=3, test_size=1):
+        self.n_splits = n_splits
+        self.test_size = test_size
+
+    def split(self, X, y=None):
+        train_size = len(X)-self.test_size
+        train_blocks = np.linspace(0, train_size, self.n_splits+1, dtype=int)
+        train_i = train_blocks[:-1]
+        train_j = train_blocks[1:]
+        for i, j in zip(train_blocks[:-1], train_blocks[1:]):
+            yield np.arange(i, j), np.arange(j, j+self.test_size)