Hotfix/cleanup

big_scape/network/families.py

@@ -4,9 +4,6 @@
 import sys
 from typing import Callable, Optional
 import warnings
-import numpy as np
-import networkx
-import math
 import logging
 
 # from dependencies
@@ -105,24 +102,6 @@ def generate_families(
     return regions_families
 
 
-def get_cc_edge_weight_std(connected_component) -> float:
-    """calculates the standard deviation of the edge weights of a connected component
-
-    Args:
-        connected_component (list[tuple[int, int, float, float, float, float, str]]):
-            connected component in the form of a list of edges
-
-    Returns:
-        float: standard deviation of the edge weights of the connected component
-    """
-
-    edge_weights = [edge[2] for edge in connected_component]
-    edge_std = np.std(edge_weights)
-    edge_std = round(edge_std, 2)
-
-    return edge_std
-
-
 def get_cc_density(
     connected_component: list[tuple[int, int, float, float, float, float, int]]
 ) -> float:
@@ -148,46 +127,6 @@ def get_cc_density(
     return cc_density
 
 
-def test_centrality(connected_component, node_fraction) -> tuple[bool, list[int]]:
-    """tests if a network will break when removing the top nodes
-    with highest betweenness centrality
-
-    Args:
-        connected_component (list[tuple[int, int, float, float, float, float, str]]):
-            connected component in the form of a list of edges
-        node_fraction (float): fraction of nodes with highest betweenness centrality to remove
-
-    Returns:
-        tuple[bool, list[int]]: whether the network breaks and the list of nodes sorted by betweenness centrality
-    """
-
-    edgelist = [(edge[0], edge[1], edge[2]) for edge in connected_component]
-
-    graph = networkx.Graph()
-    graph.add_weighted_edges_from(edgelist)
-
-    betweeness_centrality_dict = networkx.betweenness_centrality(graph)
-    sorted_between_bentrality_nodes = sorted(
-        betweeness_centrality_dict, key=betweeness_centrality_dict.get, reverse=True
-    )
-
-    # round up to nearest integer
-    top_nodes = math.ceil(len(sorted_between_bentrality_nodes) * node_fraction)
-    nodes_to_remove = sorted_between_bentrality_nodes[:top_nodes]
-
-    for node in nodes_to_remove:
-        graph.remove_node(node)
-
-    nr_ccs = networkx.number_connected_components(graph)
-
-    del graph
-
-    if nr_ccs > 1:
-        return True, sorted_between_bentrality_nodes
-
-    return False, sorted_between_bentrality_nodes
-
-
 def aff_sim_matrix(matrix, preference: Optional[float] = None):
     """Execute affinity propagation on a __similarity__ matrix
 

big_scape/network/utility.py

@@ -2,23 +2,6 @@
 
 # from dependencies
 import numpy as np
-import networkx as nx
-
-
-def sim_matrix_from_graph(graph: nx.Graph, edge_property: str) -> np.ndarray:
-    """Return a similarity matrix from a graph in the form of a numpy array
-
-    Args:
-        graph (Graph): graph
-        edge_property (str): _description_
-
-    Returns:
-        ndarray: _description_
-    """
-    matrix = nx.to_numpy_array(graph, weight=edge_property, nonedge=1.0)
-    # have to convert from distances to similarity
-    matrix = 1 - matrix
-    return matrix
 
 
 def edge_list_to_adj_list(

big_scape/utility/version.py

@@ -1,8 +1,7 @@
 """Module that contains helper functions specifically related to the bigscape version
 """
 
-import toml
-
+import tomllib
 from importlib import metadata
 from pathlib import Path
 
@@ -21,7 +20,8 @@ def get_bigscape_version() -> str:
     pyproject_toml = Path(__file__).parent.parent.parent / "pyproject.toml"
 
     if pyproject_toml.exists():
-        return toml.load(pyproject_toml)["project"]["version"]
+        with open(pyproject_toml, "rb") as fp:
+            return tomllib.load(fp)["project"]["version"]
 
     # if not, we're probably running as a package. get the version of the currently
     # installed big-scape package

environment.yml

@@ -7,7 +7,6 @@ dependencies:
   - biopython=1.81
   - sortedcontainers=2.4.0
   - fasttree=2.1.11
-  - networkx=3.1
   - numpy=1.26.0
   - pandas=2.1.1
   - pyhmmer=0.10.14

pyproject.toml

@@ -35,7 +35,6 @@ dev = [
 
     # type stubs (https://mypy.readthedocs.io/en/stable/running_mypy.html#missing-imports)
     "types-psutil",
-    "networkx-stubs",
     "data-science-types",
     "types-tqdm",
     "types-setuptools"

test/comparison/__init__.py

@@ -0,0 +1 @@
+"""Contains tests for distance calculation"""

test/hmm/__init__.py

@@ -0,0 +1 @@
+"""Contains tests involving (py)HMMer analysis and processing"""

test/network/__init__.py

@@ -0,0 +1 @@
+"""Contains tests involving CC and family generation"""

test/network/test_family.py

@@ -121,18 +121,6 @@ def test_aff_sim_matrix(self):
 
         self.assertListEqual(expected_labels, actual_labels)
 
-    def test_get_cc_edge_weight_std(self):
-        """Tests whether the standard deviation of the edge weights of a connected
-        component is correctly calculated
-        """
-        adj_list = TestAffinityPropagation.gen_edge_list()
-
-        expected_std = 0.12
-
-        actual_std = bs_families.get_cc_edge_weight_std(adj_list)
-
-        self.assertEqual(expected_std, actual_std)
-
     def test_get_cc_density(self):
         """Tests whether the density of a connected component is correctly
         calculated
@@ -146,17 +134,3 @@ def test_get_cc_density(self):
         actual_density = bs_families.get_cc_density(adj_list)
 
         self.assertEqual(expected_density, actual_density)
-
-    def test_test_centrality(self):
-        """Tests whether the test_centrality function correctly identifies a network
-        that will break when removing the top nodes with highest betweenness centrality
-        """
-        adj_list = TestAffinityPropagation.gen_edge_list_alt()
-
-        expected_break = True
-
-        actual_break, actual_sorted_centrality_nodes = bs_families.test_centrality(
-            adj_list, 0.3
-        )
-
-        self.assertEqual(expected_break, actual_break)