Fix deprecated NumPy usage and updated test data

netomaton/evolution.py

@@ -418,7 +418,7 @@ def _get_input_function(timesteps=None, input=None):
 
 
 def _is_indexable(obj):
-    return isinstance(obj, collections.Sequence)
+    return isinstance(obj, collections.abc.Sequence)
 
 
 class _TimestepInputFunction:
@@ -441,7 +441,7 @@ def __call__(self, t, activities, network):
         return self._input_list[t-1]
 
 
-def init_simple(size, val=1, dtype=np.int):
+def init_simple(size, val=1, dtype=int):
     """
     Returns an array initialized with zeroes, with its center value set to the specified value, or 1 by default.
     :param size: the size of the array to be created
@@ -454,7 +454,7 @@ def init_simple(size, val=1, dtype=np.int):
     return x.tolist()
 
 
-def init_random(size, k=2, n_randomized=None, empty_value=0, dtype=np.int):
+def init_random(size, k=2, n_randomized=None, empty_value=0, dtype=int):
     """
     Returns a randomly initialized array with values consisting of numbers in {0,...,k - 1}, where k = 2 by default.
     If dtype is not an integer type, then values will be uniformly distributed over the half-open interval [0, k - 1).
@@ -480,7 +480,7 @@ def init_random(size, k=2, n_randomized=None, empty_value=0, dtype=np.int):
     return np.array(np.pad(np.array(rand_nums), (pad_left, pad_right), 'constant', constant_values=empty_value)).tolist()
 
 
-def init_simple2d(rows, cols, val=1, dtype=np.int):
+def init_simple2d(rows, cols, val=1, dtype=int):
     """
     Returns a list initialized with zeroes, with its center value set to the specified value, or 1 by default, when the
     list is reshaped according to the given number of rows and columns.
@@ -497,5 +497,5 @@ def init_simple2d(rows, cols, val=1, dtype=np.int):
 
 def check_np(obj):
     if isinstance(obj, np.generic):
-        return np.asscalar(obj)
+        return obj.item()
     return obj

netomaton/hopfield_net.py

@@ -21,7 +21,7 @@ def train(self, P):
         adjacency matrix. Therefore, before the network can be used to evaluate a pattern, it must be trained.
         :param P: the set of training patterns, as a list of lists
         """
-        self._adjacency_matrix = np.zeros((len(P[0]), len(P[0])), dtype=np.int)
+        self._adjacency_matrix = np.zeros((len(P[0]), len(P[0])), dtype=int)
         for p in P:
             for i in range(len(p)):
                 for j in range(len(p)):

netomaton/langtons_loop.py

@@ -251,7 +251,7 @@ def init_loops(self, n, row, col):
         :return: the initial conditions
         """
 
-        initial_conditions = np.zeros(self._dim, dtype=np.int)
+        initial_conditions = np.zeros(self._dim, dtype=int)
         for i in range(n):
             row_i = row[i]
             col_i = col[i]

netomaton/rule_tables.py

@@ -44,7 +44,7 @@ def random_rule_table(k, r, lambda_val=None, quiescent_state=None, strong_quiesc
     if lambda_val is None:
         lambda_val = 1. - (1. / k)
     if quiescent_state is None:
-        quiescent_state = np.random.randint(k, dtype=np.int)
+        quiescent_state = np.random.randint(k, dtype=int)
     if not (0 <= quiescent_state <= k - 1):
         raise Exception("quiescent state must be a number in {0,...,k - 1}")
     other_states = [x for x in range(0, k) if x != quiescent_state]

pyproject.toml

@@ -0,0 +1,3 @@
+[build-system]
+requires = ["setuptools>=64.0"]
+build-backend = "setuptools.build_meta"

tests/rule_test.py

@@ -18,7 +18,7 @@ def _convert_to_matrix(self, filename, dtype=int):
         content = content.replace('{', '')
         content = content.replace('},', ';')
         content = [[dtype(i) for i in x.split(',')] for x in content.split(';')]
-        return np.array(content, dtype=np.int)
+        return np.array(content, dtype=int)
 
     def _convert_to_matrix2d(self, filename, dtype=int):
         with open(os.path.join(THIS_DIR, 'resources', filename), 'r') as content_file:
@@ -30,7 +30,7 @@ def _convert_to_matrix2d(self, filename, dtype=int):
         content = [x.split('},') for x in content.split('}},')]
         content = [[h.split(',') for h in x] for x in content]
         content = [[[dtype(i) for i in h] for h in x] for x in content]
-        return np.array(content, dtype=np.int)
+        return np.array(content, dtype=int)
 
     def _convert_to_list_of_lists(self, filename, strings=False, dtype=int):
         with open(os.path.join(THIS_DIR, 'resources', filename), 'r') as content_file: