Clean up unused imports (#319)

See #318. This gets almost all the LGTM issues.

netrd/distance/dmeasure.py

@@ -109,9 +109,6 @@ def dist(self, G1, G2, w1=0.45, w2=0.45, w3=0.10, niter=50):
         second_term = 0
         third_term = 0
 
-        N1 = G1.number_of_nodes()
-        N2 = G2.number_of_nodes()
-
         if w1 + w2 > 0:
             g1_nnd, g1_pdfs = network_node_dispersion(G1)
             g2_nnd, g2_pdfs = network_node_dispersion(G2)

netrd/distance/ipsen_mikhailov.py

@@ -58,8 +58,6 @@ def dist(self, G1, G2, hwhm=0.08):
         .. [1] https://journals.aps.org/pre/abstract/10.1103/PhysRevE.66.046109
 
         """
-        N = len(G1)
-
         # get the adjacency matrices
         adj1 = nx.to_numpy_array(G1)
         adj2 = nx.to_numpy_array(G2)

netrd/distance/nbd.py

@@ -384,8 +384,4 @@ def euclidean_distance(x, y):
     if status not in [solver.OPTIMAL, solver.FEASIBLE]:
         raise Exception("Unable to find feasible solution")
 
-    for ((x, y), variable) in variables.items():
-        if variable.solution_value() != 0:
-            cost = euclidean_distance(x, y) * variable.solution_value()
-
     return objective.Value()

netrd/distance/netsimile.py

@@ -92,7 +92,6 @@ def feature_extraction(G):
     node_degree_dict = dict(G.degree())
     node_clustering_dict = dict(nx.clustering(G))
     egonets = [nx.ego_graph(G, n) for n in node_list]
-    egonet_dict = {node: egonet for node, egonet in zip(node_list, egonets)}
 
     # node degrees
     degs = [node_degree_dict[n] for n in node_list]

netrd/distance/onion_divergence.py

@@ -132,7 +132,6 @@ def _onion_decomposition(G):
 def _update_sparse_matrix(dictionary, values, key, index):
     # Creates entry if it doesn't exist or else update the corresponding value
     if not key in dictionary:
-        index_i = index
         dictionary[key] = index
         values.append(1)
         index += 1

netrd/distance/portrait_divergence.py

@@ -39,7 +39,7 @@
 from collections import Counter
 import numpy as np
 import networkx as nx
-from ..utilities import entropy, unweighted
+from ..utilities import entropy
 
 
 class PortraitDivergence(BaseDistance):

netrd/dynamics/branching_process.py

@@ -20,9 +20,6 @@
 class BranchingModel(BaseDynamics):
     """A sand-pile-like branching process."""
 
-    def __init__(self):
-        self.results = {}
-
     def simulate(
         self,
         G,
@@ -115,7 +112,6 @@ def simulate(
                 active_edges = G.out_edges(nbunch=active_nodes, data=True)
 
                 if len(active_edges) != 0:
-                    current_sources = list(list(zip(*active_edges))[0])
                     current_targets = list(list(zip(*active_edges))[1])
                     weights_array = np.array([j[2]['weight'] for j in active_edges])
 
@@ -174,14 +170,16 @@ def initialize_history(N, L, initial_fraction, m, target_Ahat, noise):
 
     # maybe also here initialize TS_init with external drives?
     if m != 1.0:
-        N_nodes = 1000
         if N > 1000:
             N_nodes = N
+        else:
+            N_nodes = 1000
         h_vals = np.random.poisson(target_Ahat * N_nodes * np.abs(1 - m), L)
     else:
-        N_nodes = 100
         if N > 100:
             N_nodes = N
+        else:
+            N_nodes = 100
         h_vals = np.random.poisson(0.01, L)
     #         h_vals = np.random.poisson(target_Ahat*N_nodes * 0.01, L)
 

netrd/reconstruction/correlation_matrix.py

@@ -8,7 +8,6 @@
 """
 from .base import BaseReconstructor
 import numpy as np
-import networkx as nx
 from ..utilities import create_graph, threshold
 
 

netrd/reconstruction/correlation_spanning_tree.py

@@ -103,8 +103,6 @@ def fit(self, TS, distance='root_inv', **kwargs):
                topology of stock markets. Quantitative Finance 1, 372–374.
 
         """
-        N, L = TS.shape
-
         C = np.corrcoef(TS)  # Empirical correlation matrix
 
         D = (

netrd/reconstruction/free_energy_minimization.py

@@ -8,8 +8,6 @@
 """
 from .base import BaseReconstructor
 import numpy as np
-import networkx as nx
-import scipy as sp
 from scipy import linalg
 from ..utilities import create_graph, threshold
 

netrd/reconstruction/maximum_likelihood_estimation.py

@@ -8,7 +8,6 @@
 """
 from .base import BaseReconstructor
 import numpy as np
-import networkx as nx
 from ..utilities import create_graph, threshold
 
 

netrd/reconstruction/mean_field.py

@@ -8,8 +8,6 @@
 """
 from .base import BaseReconstructor
 import numpy as np
-import networkx as nx
-import scipy as sp
 from scipy import linalg
 from scipy.integrate import quad
 from scipy.optimize import fsolve

netrd/reconstruction/mutual_information_matrix.py

@@ -14,7 +14,6 @@
 
 from .base import BaseReconstructor
 import numpy as np
-import networkx as nx
 from ..utilities import create_graph, threshold
 
 

netrd/reconstruction/ou_inference.py

@@ -14,9 +14,8 @@
 """
 
 from .base import BaseReconstructor
-import networkx as nx
 import numpy as np
-from scipy.linalg import eig, inv
+from scipy.linalg import eig
 from ..utilities import create_graph, threshold
 
 

netrd/reconstruction/partial_correlation_matrix.py

@@ -11,7 +11,6 @@
 """
 from .base import BaseReconstructor
 import numpy as np
-import networkx as nx
 from scipy import stats, linalg
 from ..utilities import create_graph, threshold
 

netrd/reconstruction/thouless_anderson_palmer.py

@@ -9,8 +9,6 @@
 """
 from .base import BaseReconstructor
 import numpy as np
-import networkx as nx
-import scipy as sp
 from scipy import linalg
 from ..utilities import create_graph, threshold
 

netrd/utilities/cluster.py

@@ -10,7 +10,6 @@
 Submitted as part of the 2019 NetSI Collabathon.
 
 """
-import numpy as np
 import networkx as nx
 from scipy.cluster.hierarchy import dendrogram, linkage
 

netrd/utilities/read.py

@@ -10,7 +10,6 @@
 
 """
 import numpy as np
-import networkx as nx
 
 
 def read_time_series(filename, delimiter=','):