Add files via upload

README.md

@@ -66,13 +66,6 @@ $ conda install -c tommasobelluzzo pydtmc
 $ conda update -c tommasobelluzzo pydtmc
 ```
 
-[GitHub](https://github.com/):
-
-```sh
-$ pip install git+https://github.com/TommasoBelluzzo/PyDTMC.git@master#egg=PyDTMC
-$ pip install --upgrade git+https://github.com/TommasoBelluzzo/PyDTMC.git@master#egg=PyDTMC
-```
-
 ## Usage
 
 The core element of the library is the `MarkovChain` class, which can be instantiated as follows:

meta.yaml

@@ -44,7 +44,7 @@ test:
 about:
   home: {{ data['url'] }}
   license: {{ data['license'] }}
-  license_file: {{ data['license_files'] }}
+  license_file: {{ data['url'] }}/blob/master/LICENSE.md
   summary: {{ summary }}
   description: {{ description }}
   dev_url: {{ data['url'] }}

pydtmc/__init__.py

@@ -1,26 +1,26 @@
-# -*- coding: utf-8 -*-
-
-__title__ = 'PyDTMC'
-__version__ = '6.1.0'
-__author__ = 'Tommaso Belluzzo'
-
-__all__ = [
-    'ValidationError',
-    'MarkovChain',
-    'plot_eigenvalues', 'plot_graph', 'plot_redistributions', 'plot_walk'
-]
-
-from pydtmc.exceptions import (
-    ValidationError
-)
-
-from pydtmc.markov_chain import (
-    MarkovChain
-)
-
-from pydtmc.plotting import (
-    plot_eigenvalues,
-    plot_graph,
-    plot_redistributions,
-    plot_walk
-)
+# -*- coding: utf-8 -*-
+
+__title__ = 'PyDTMC'
+__version__ = '6.1.0'
+__author__ = 'Tommaso Belluzzo'
+
+__all__ = [
+    'ValidationError',
+    'MarkovChain',
+    'plot_eigenvalues', 'plot_graph', 'plot_redistributions', 'plot_walk'
+]
+
+from pydtmc.exceptions import (
+    ValidationError
+)
+
+from pydtmc.markov_chain import (
+    MarkovChain
+)
+
+from pydtmc.plotting import (
+    plot_eigenvalues,
+    plot_graph,
+    plot_redistributions,
+    plot_walk
+)

pydtmc/markov_chain.py

@@ -1959,15 +1959,21 @@ def from_graph(graph: tgraphs) -> tmc:
         size = len(states)
 
         p = np.zeros((size, size), dtype=float)
+        edges = list(graph.edges(data='weight', default=0.0))
 
-        for state_from, weights in graph.adjacency():
+        for edge in edges:
+            i = states.index(edge[0])
+            j = states.index(edge[1])
+            p[i, j] = float(edge[2])
 
-            i = states.index(state_from)
+        p_sums = np.sum(p, axis=1)
 
-            for state_to, data in weights.items():
-                j = states.index(state_to)
-                w = data['weight']
-                p[i, j] = w
+        for i in range(size):
+
+            if np.isclose(p_sums[i], 0.0):  # pragma: no cover
+                p[i, :] = np.ones(p.shape[0], dtype=float) / size
+            else:
+                p[i, :] /= p_sums[i]
 
         mc = MarkovChain(p, states)
 
@@ -2069,13 +2075,14 @@ def from_matrix(m: tnumeric, states: olist_str = None) -> tmc:
             raise generate_validation_error(e, trace()) from None
 
         p = np.copy(m)
-        p_size = p.shape[0]
         p_sums = np.sum(p, axis=1)
 
-        for i in range(p_size):
+        size = p.shape[0]
+
+        for i in range(size):
 
             if np.isclose(p_sums[i], 0.0):  # pragma: no cover
-                p[i, :] = np.ones(p.shape[0], dtype=float) / p_size
+                p[i, :] = np.ones(p.shape[0], dtype=float) / size
             else:
                 p[i, :] /= p_sums[i]