finish refactoring

dominoes.ipynb

@@ -55,14 +55,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 1,
+   "execution_count": 30,
    "id": "40c715a5",
    "metadata": {},
    "outputs": [],
    "source": [
     "# TODO For refactoring\n",
     "# continue working on generate_batch method in DominoesDataset\n",
-    "# "
+    "# generate_batch method needs help with \"set_target\""
    ]
   },
   {

dominoes/utils.py

@@ -94,85 +94,6 @@ def listDominoes(highestDominoe):
     return get_dominoe_set(highestDominoe, as_torch=False)
 
 
-def get_dominoe_set(highest_dominoe, as_torch=False):
-    """
-    Create a list of dominoes in a set with highest value of <highest_dominoe>
-
-    The dominoes are paired values (combinations with replacement) of integers
-    from 0 to <highest_dominoe>. This method returns either a numpy array or a
-    torch tensor of the dominoes as integers.
-
-    The shape will be (num_dominoes, 2) where the first column is the first value
-    of the dominoe, and the second column is the second value of the dominoe.
-
-    args:
-        highest_dominoe: the highest value of a dominoe
-        as_torch: return dominoes as torch tensor if True, otherwise return numpy array
-
-    returns:
-        dominoes: an array or tensor of dominoes in the set
-    """
-    # given a standard rule for how to organize the list of dominoes as one-hot arrays, list the dominoes present in a one hot array
-    array_function = torch.tensor if as_torch else np.array
-    stack_function = torch.stack if as_torch else np.stack
-    dominoe_set = [array_function(quake, dtype=int) for quake in itertools.combinations_with_replacement(np.arange(highest_dominoe + 1), 2)]
-    return stack_function(dominoe_set)
-
-
-def twohot_dominoe(selected, dominoes, highest_dominoe, available=None, available_token=False, null_token=False, with_batch=True):
-    """
-    converts an index of selected dominoes to a stacked two-hot representation
-
-    dominoes are paired values (combinations with replacement) of integers
-    from 0 to <highest_dominoe>.
-
-    This simple representation is a two-hot vector where the first
-    <highest_dominoe>+1 elements represent the first value of the dominoe, and
-    the second <highest_dominoe>+1 elements represent the second value of the
-    dominoe. Here are some examples for <highest_dominoe> = 3:
-
-    (0 | 0): [1, 0, 0, 0, 1, 0, 0, 0]
-    (0 | 1): [1, 0, 0, 0, 0, 1, 0, 0]
-    (0 | 2): [1, 0, 0, 0, 0, 0, 1, 0]
-    (0 | 3): [1, 0, 0, 0, 0, 0, 0, 1]
-    (1 | 0): [0, 1, 0, 0, 1, 0, 0, 0]
-    (2 | 1): [0, 0, 1, 0, 0, 1, 0, 0]
-
-    """
-    assert selected.ndim == 1, "selected must have shape (num_selected, )"
-    if available_token:
-        assert available is not None, "if with_available=True, then available needs to be provided"
-    num_dominoes = selected.shape[0]
-
-    # input dimension determined by highest dominoe (twice the number of possible values on a dominoe)
-    input_dim = (2 if not (available_token) else 3) * (highest_dominoe + 1) + (1 if null_token else 0)
-
-    # first & second value are index and shifted index
-    firstValue = torch.tensor(dominoes[selected, 0], dtype=torch.int64).unsqueeze(1)
-    secondValue = torch.tensor(dominoes[selected, 1] + highest_dominoe + 1, dtype=torch.int64).unsqueeze(1)
-
-    # scatter data into two-hot vectors
-    src = torch.ones((num_dominoes, 1), dtype=torch.float)
-    twohot = torch.zeros((num_dominoes, input_dim), dtype=torch.float)
-    twohot.scatter_(1, firstValue, src)
-    twohot.scatter_(1, secondValue, src)
-
-    if null_token:
-        null = torch.zeros((1, input_dim), dtype=torch.float).scatter_(1, torch.tensor(input_dim - 1).view(1, 1), torch.tensor(1.0).view(1, 1))
-        twohot = torch.cat((twohot, null), dim=0)
-
-    if available_token:
-        rep_available = torch.zeros((1, input_dim), dtype=torch.float)
-        availableidx = int((highest_dominoe + 1) * 2 + available)
-        rep_available[0, availableidx] = 1.0
-        twohot = torch.cat((twohot, rep_available), dim=0)
-
-    if with_batch:
-        twohot = twohot.unsqueeze(0)
-
-    return twohot
-
-
 def dominoesString(dominoe):
     return f"{dominoe[0]:>2}|{dominoe[1]:<2}"