Numpy check

edflow/data/dataset_mixin.py

@@ -1,6 +1,6 @@
 from chainer.dataset import DatasetMixin as DatasetMixin_
 import numpy as np
-from edflow.util import walk, update
+from edflow.util import walk, update, LabelsDict
 
 # handle bug with mocked chainer.dataset.DatasetMixin import
 if hasattr(DatasetMixin_, "_mock_name"):
@@ -287,6 +287,7 @@ def labels(self):
 
     @labels.setter
     def labels(self, labels):
+        labels = LabelsDict(labels)
         if hasattr(self, "data"):
             self.data.labels = labels
         else:

edflow/edsetup_files/dataset.py

@@ -16,7 +16,7 @@ def __init__(self, config):
         self.example_names = config.get("example_names")
 
         self.inputs = np.random.rand(self.num_example, self.feature_dimension)
-        self.labels = self.inputs
+        self.labels = {"random": self.inputs}
 
     def get_example(self, idx):
         """
@@ -31,7 +31,7 @@ def get_example(self, idx):
         example (dict): These will be retrieved by their respective keys in the step_op method of the iterator.
         """
         inputs = self.inputs[idx]
-        labels = self.labels[idx]
+        labels = self.labels["random"][idx]
 
         example = {"inputs": inputs, "labels": labels}
         return example

edflow/util.py

@@ -829,6 +829,45 @@ def __init__(self, config):
         self.config = config
 
 
+class LabelsDict(dict):
+    """
+    Dict which asserts all values are numpy arrays or memmaps.
+    Code from https://stackoverflow.com/questions/2060972/subclassing-python-dictionary-to-override-setitemZ
+    """
+
+    def __init__(self, *args, **kwargs):
+        self.update(*args, **kwargs)
+
+    def __setitem__(self, key, value):
+        # optional processing here
+        def assert_valid_label(val):
+            assert isinstance(
+                val, (np.ndarray, np.memmap)
+            ), "Labels must be dict s of numpy arrays or memmaps and not {}!".format(
+                type(val)
+            )
+
+        walk(value, assert_valid_label)
+        super(LabelsDict, self).__setitem__(key, value)
+
+    def update(self, *args, **kwargs):
+        if args:
+            if len(args) > 1:
+                raise TypeError(
+                    "update expected at most 1 arguments, " "got %d" % len(args)
+                )
+            other = dict(args[0])
+            for key in other:
+                self[key] = other[key]
+        for key in kwargs:
+            self[key] = kwargs[key]
+
+    def setdefault(self, key, value=None):
+        if key not in self:
+            self[key] = value
+        return self[key]
+
+
 if __name__ == "__main__":
     from edflow.data.util import plot_datum
 

setup.py

@@ -30,7 +30,7 @@
     extras_require={
         "explore": ["streamlit > 0.49"],
         "docs": ["sphinx >= 1.4", "sphinx_rtd_theme", "better-apidoc"],
-        "test": ["pytest", "pytest-cov", "coveralls"],
+        "test": ["pytest", "pytest-cov", "coveralls", "coverage < 5.0"],
     },
     zip_safe=False,
     scripts=[

tests/test_data/test_datasetmixin.py

@@ -158,9 +158,34 @@ def __len__(self):
     assert d2["val"] == d1["val"]()
 
 
+def test_labels_numpy():
+    class MyDset(DatasetMixin):
+        def __init__(self):
+            self.labels = {}
+
+    # set labels with array
+    dataset = MyDset()
+    dataset.labels = {"l": np.arange(10)}
+
+    # set labels with list
+    dataset = MyDset()
+    with pytest.raises(AssertionError):
+        dataset.labels = {"l": list(np.arange(10))}
+
+    # update labels with array
+    dataset = MyDset()
+    dataset.labels.update({"l2": np.arange(10)})
+
+    # update labels with list
+    dataset = MyDset()
+    with pytest.raises(AssertionError):
+        dataset.labels.update({"l2": list(np.arange(10))})
+
+
 if __name__ == "__main__":
     test_dset_mxin()
     test_dset_mxin_ops()
     test_dset_mxin_data_attr()
     test_dset_mxin_app_labels()
     test_dset_mxin_data_attr_app_labels()
+    test_labels_numpy()

tests/test_util.py

@@ -8,9 +8,11 @@
     contains_key,
     KeyNotFoundError,
     get_leaf_names,
+    LabelsDict,
 )
 from edflow import util
 from itertools import product
+import numpy as np
 
 # ================= set_value ====================
 
@@ -792,3 +794,25 @@ def test_get_leaf_name():
     ref = sorted(["a/0", "a/1", "b/c/d", "e"])
 
     assert names == ref
+
+
+# ================= LabelsDict ===================
+
+
+def test_LabelsDict():
+    """Code adapted from https://stackoverflow.com/questions/2060972/subclassing-python-dictionary-to-override-setitem"""
+
+    def test_updates(dictish):
+        dictish["abc"] = np.array(123)
+        dictish.update({"def": np.array(234)})
+        dictish.update(red=np.array(1), blue=np.array(2))
+        dictish.update([("orange", np.array(3)), ("green", np.array(4))])
+        dictish.update({"hello": np.array("kitty")}, black=np.array("white"))
+        dictish.update({"yellow": np.array(5)}, yellow=np.array(6))
+        dictish.setdefault("brown", np.array(7))
+
+    python_dict = dict([("b", 2), ("c", 3)], a=1)
+    test_updates(python_dict)
+
+    my_dict = LabelsDict([("b", np.array(2)), ("c", np.array(3))], a=np.array(1))
+    test_updates(my_dict)