Merge pull request #36 from jussiviinikka/feature/param-refactoring-a…

…nd-adjustment

Feature/param refactoring and adjustment

.git-blame-ignore-revs

@@ -13,3 +13,6 @@ dc29f6da14c3a11e9f523001ef7c79be8832ad9c
 
 # PR 11: Use isort (and a bit of black)
 8fe07a70bd5dd104bee44405cf9a35ec388420a5
+
+# Use tabs for c++ indentation consistently
+c46d5467b3109c93a103795080a20416f7fdb0a9

setup.cfg

@@ -5,9 +5,16 @@ version = attr: sumu.__version__
 max-line-length = 79
 target-version = ['py37']
 ignore=
-    E731,  # do not assign a lambda expression, use a def
-    E203,  # space before : (needed for how black formats slicing)
-	W605,  # invalid escape sequence '\m' (m is any letter) :
-	       # (needed for TeX)
-    W503,  # line break before binary operator
-    E741,  # do not use variables named ‘l’, ‘O’, or ‘I’
+    # do not assign a lambda expression, use a def
+    E731,
+    # space before : (needed for how black formats slicing)
+    E203,
+    # invalid escape sequence '\m' (m is any letter) :
+    # (needed for TeX)
+    W605,
+    # line break before binary operator
+    W503,
+    # line break after binary operator
+    W504,
+    # do not use variables named ‘l’, ‘O’, or ‘I’
+    E741,

setup.py

@@ -187,10 +187,10 @@ def run(self):
     author_email="jussi.viinikka@helsinki.fi",
     license="BSD",
     packages=["sumu", "sumu.utils", "sumu.scores"],
-    install_requires=["numpy", "scipy>=1.6"],
+    install_requires=["numpy", "scipy>=1.6", "psutil"],
     package_data={"sumu": ["sources.bib"]},
     include_package_data=True,
-    extras_require={"plotext": ["plotext==4.1.3"], "test": ["psutil"]},
+    extras_require={"plotext": ["plotext==4.1.3"]},
     cmdclass=cmdclass,
     ext_modules=cythonize(
         exts, language_level="3", compiler_directives=COMPILER_DIRECTIVES

sumu/__init__.py

@@ -1,4 +1,4 @@
-__version__ = "0.2.dev5"
+__version__ = "0.2.dev6"
 
 from . import bnet, gadget
 from .aps import aps

sumu/_mcmc_moves.pyx

@@ -6,7 +6,7 @@ from .bnet import transitive_closure
 from .utils.math_utils import comb, subsets
 
 
-def DAG_edgerev(**kwargs):
+def DAG_edge_reversal(**kwargs):
 
     DAG = kwargs["DAG"]
     R = kwargs["R"]
@@ -77,7 +77,7 @@ def DAG_edgerev(**kwargs):
     return DAG, scoreratio, edge
 
 
-def R_basic_move(**kwargs):
+def R_split_merge(**kwargs):
     """Splits or merges a root-partition :footcite:`kuipers:2017`.
 
     Args:
@@ -138,7 +138,7 @@ def R_basic_move(**kwargs):
     return R_prime, q, q_prime, rescore
 
 
-def R_swap_any(**kwargs):
+def R_swap_node_pair(**kwargs):
 
     R = kwargs["R"]
     m = len(R)

sumu/candidates.py

@@ -53,18 +53,18 @@ def greedy(
     K,
     *,
     scores,
-    params={"k": 6, "t_budget": None, "criterion": "score"},
+    params={"K_f": 6, "t_budget": None, "association_measure": "score"},
     **kwargs,
 ):
-    k = params.get("k")
+    k = params.get("K_f")
     if k is not None:
         k = min(k, K)
     t_budget = params.get("t_budget")
-    criterion = params.get("criterion")
-    assert criterion in ("score", "gain")
-    if criterion == "score":
+    association_measure = params.get("association_measure")
+    assert association_measure in ("score", "gain")
+    if association_measure == "score":
         goodness = lambda v, S, u: scores._local(v, np.array(S + (u,)))
-    elif criterion == "gain":
+    elif association_measure == "gain":
         goodness = lambda v, S, u: scores._local(
             v, np.array(S + (u,))
         ) - scores._local(v, np.array(S))
@@ -136,7 +136,7 @@ def get_k(t_budget):
         C[v] = tuple(C[v])
         scores.clear_cache()
 
-    return C, {"k": k}
+    return C, {"K_f": k}
 
 
-candidate_parent_algorithm = {"opt": opt, "rnd": rnd, "greedy": greedy}
+candidate_parent_algorithm = {"optimal": opt, "random": rnd, "greedy": greedy}

sumu/data.py

@@ -67,6 +67,10 @@ def n(self):
     def N(self):
         return self.data.shape[0]
 
+    @property
+    def shape(self):
+        return self.data.shape
+
     @property
     def arities(self):
         return np.count_nonzero(np.diff(np.sort(self.data.T)), axis=1) + 1