Feat: Make some deps optional (#90)

setup.py

@@ -11,6 +11,7 @@
         ["src/pygenstability/generalized_louvain/generalized_louvain.cpp"],
         include_dirs=["extra", "generalizedLouvain"],
         extra_compile_args=["-std=c++11"],
+        optional=True,
     ),
 ]
 plotly_require = ["plotly>=3.6.0"]
@@ -28,17 +29,17 @@
     "numpy>=1.18.1",
     "scipy>=1.4.1",
     "matplotlib>=3.1.3",
-    "networkx>=3.0",
     "scikit-learn",
     "cmake>=3.16.3",
     "click>=7.0",
     "tqdm>=4.45.0",
     "pybind11>=2.10.0",
     "pandas>=1.0.0",
-    "igraph",
-    "leidenalg",
     "threadpoolctl",
 ]
+leiden_install = ["igraph", "leidenalg"]
+networkx_install = ["networkx>=3.0"]
+
 setup(
     name="PyGenStability",
     version=__version__,
@@ -54,7 +55,9 @@
     zip_safe=False,
     extras_require={
         "plotly": plotly_require,
-        "all": plotly_require + test_require,
+        "leiden": leiden_install,
+        "networkx": networkx_install,
+        "all": plotly_require + test_require + leiden_install + networkx_install,
     },
     entry_points={"console_scripts": ["pygenstability=pygenstability.app:cli"]},
     packages=find_namespace_packages("src"),

src/pygenstability/plotting.py

@@ -4,7 +4,12 @@
 
 import matplotlib
 import matplotlib.pyplot as plt
-import networkx as nx
+
+try:
+    import networkx as nx
+except ImportError:  # pragma: no cover
+    print('Please install networkx via pip install "pygenstability[networkx]" for full plotting.')
+
 import numpy as np
 from matplotlib import gridspec
 from matplotlib import patches

src/pygenstability/pygenstability.py

@@ -19,8 +19,14 @@
 from functools import wraps
 from time import time
 
-import igraph as ig
-import leidenalg
+try:
+    import igraph as ig
+    import leidenalg
+
+    _NO_LEIDEN = False
+except ImportError:  # pragma: no cover
+    _NO_LEIDEN = True
+
 import numpy as np
 import scipy.sparse as sp
 from sklearn.metrics import mutual_info_score
@@ -101,6 +107,21 @@ def _get_constructor_data(constructor, scales, pool, tqdm_disable=False):
     )
 
 
+def _check_method(method):  # pragma: no cover
+    if _NO_LEIDEN and not hasattr(generalized_louvain, "evaluate_quality"):
+        raise Exception("Without Louvain or Leiden solver, we cannot run PyGenStability")
+
+    if method == "louvain" and not hasattr(generalized_louvain, "evaluate_quality"):
+        print("Louvain is not available, we fallback to leiden.")
+        return "leiden"
+
+    if method == "leiden" and _NO_LEIDEN:
+        print("Leiden is not available, we fallback to louvain.")
+        return "louvain"
+
+    return method
+
+
 @_timing
 def run(
     graph=None,
@@ -171,6 +192,7 @@ def run(
             - 'NVI': NVI at each scale
             - 'ttprime': ttprime matrix
     """
+    method = _check_method(method)
     run_params = _get_params(locals())
     graph = _graph_checks(graph)
     scales = _get_scales(
@@ -217,7 +239,7 @@ def run(
 
         if with_postprocessing:
             L.info("Apply postprocessing...")
-            _apply_postprocessing(all_results, pool, constructor_data, tqdm_disable)
+            _apply_postprocessing(all_results, pool, constructor_data, tqdm_disable, method=method)
 
         if with_ttprime or with_optimal_scales:
             L.info("Compute ttprimes...")
@@ -342,18 +364,34 @@ def _optimise(_, quality_indices, quality_values, null_model, global_shift, meth
     return stability + global_shift, community_id
 
 
-def _evaluate_quality(partition_id, qualities_index, null_model, global_shift):
+def _evaluate_quality(partition_id, qualities_index, null_model, global_shift, method="louvain"):
     """Worker for generalized Markov Stability optimisation runs."""
-    quality = generalized_louvain.evaluate_quality(
-        qualities_index[0][0],
-        qualities_index[0][1],
-        qualities_index[1],
-        len(qualities_index[1]),
-        null_model,
-        np.shape(null_model)[0],
-        1.0,
-        partition_id,
-    )
+    if method == "louvain":
+        quality = generalized_louvain.evaluate_quality(
+            qualities_index[0][0],
+            qualities_index[0][1],
+            qualities_index[1],
+            len(qualities_index[1]),
+            null_model,
+            np.shape(null_model)[0],
+            1.0,
+            partition_id,
+        )
+
+    if method == "leiden":
+        quality = np.mean(
+            [
+                leidenalg.CPMVertexPartition(
+                    ig.Graph(edges=zip(*qualities_index[0]), directed=True),
+                    initial_membership=partition_id,
+                    weights=qualities_index[1],
+                    node_sizes=null.tolist(),
+                    correct_self_loops=True,
+                ).quality()
+                for null in null_model[::2]
+            ]
+        )
+
     return quality + global_shift
 
 
@@ -390,7 +428,7 @@ def _compute_ttprime(all_results, pool):
 
 
 @_timing
-def _apply_postprocessing(all_results, pool, constructors, tqdm_disable=False):
+def _apply_postprocessing(all_results, pool, constructors, tqdm_disable=False, method="louvain"):
     """Apply postprocessing."""
     all_results_raw = all_results.copy()
 
@@ -402,6 +440,7 @@ def _apply_postprocessing(all_results, pool, constructors, tqdm_disable=False):
             qualities_index=_to_indices(constructor["quality"]),
             null_model=constructor["null_model"],
             global_shift=constructor.get("shift", 0.0),
+            method=method,
         )
         best_quality_id = np.argmax(
             pool.map(

tests/data/test_run_default_leiden.yaml

@@ -43,28 +43,28 @@ community_id:
   - 18.0
   - 19.0
   - 20.0
-- - 0.0
-  - 0.0
-  - 0.0
-  - 0.0
-  - 0.0
-  - 0.0
-  - 0.0
-  - 0.0
+- - 1.0
+  - 2.0
+  - 3.0
+  - 4.0
+  - 5.0
+  - 6.0
+  - 7.0
+  - 8.0
+  - 9.0
+  - 10.0
   - 0.0
   - 0.0
-  - 2.0
-  - 2.0
-  - 1.0
-  - 1.0
-  - 1.0
-  - 1.0
-  - 1.0
-  - 1.0
-  - 1.0
-  - 1.0
-  - 1.0
-  - 1.0
+  - 11.0
+  - 12.0
+  - 13.0
+  - 14.0
+  - 15.0
+  - 16.0
+  - 17.0
+  - 18.0
+  - 19.0
+  - 20.0
 - - 0.0
   - 0.0
   - 0.0
@@ -90,8 +90,8 @@ community_id:
 number_of_communities:
 - 21.0
 - 21.0
-- 20.0
-- 3.0
+- 21.0
+- 21.0
 - 3.0
 run_params:
   constructor: linearized
@@ -123,6 +123,6 @@ scales:
 stability:
 - 0.8526881720430108
 - 0.8526881720430108
-- 0.6401045688749567
-- 0.5087182926944845
-- 0.5087182926944845
+- 0.8526881720430108
+- 0.8526881720430108
+- 0.49930627818244916

tests/test_pygenstability.py

@@ -168,3 +168,8 @@ def test__evaluate_quality(graph):
     qualities_index = pgs._to_indices(data["quality"])
     quality = pgs._evaluate_quality(community_id, qualities_index, data["null_model"], 0)
     assert_almost_equal(quality, 0.5590341906608186)
+
+    quality = pgs._evaluate_quality(
+        community_id, qualities_index, data["null_model"], 0, method="leiden"
+    )
+    assert_almost_equal(quality, 0.2741359784037568)