chore: Sync the current main branch (#35)

* feat: add get_common_ancestor between two nodes

Co-authored-by: molkjar <[email protected]>

* bump numpy version; add integer conversion in get_common_ancestor for numba

* chore: Update dependencies to reflect main

---------

Co-authored-by: Neclow <[email protected]>
Co-authored-by: molkjar <[email protected]>

Cargo.toml

@@ -8,7 +8,7 @@ default-members = ["phylo2vec"]
 
 [workspace.package]
 edition = "2021"
-version = "0.1.0"
+version = "0.2.0"
 authors = ["Neil Scheidwasser <[email protected]>"]
 description = "Phylo2Vec: integer vector representation of binary (phylogenetic) trees"
 license = "LGPL-3.0"

py-phylo2vec/phylo2vec/tests/test_utils.py

@@ -14,9 +14,10 @@
     apply_label_mapping,
     create_label_mapping,
     check_v,
+    find_num_leaves,
+    get_common_ancestor,
     remove_leaf,
     sample,
-    find_num_leaves,
 )
 
 
@@ -103,5 +104,23 @@ def test_remove_and_add(n_leaves):
         assert np.array_equal(v, v_add)
 
 
+@pytest.mark.parametrize("n_leaves", range(MIN_N_LEAVES, MAX_N_LEAVES + 1))
+def test_get_common_ancestor(n_leaves):
+    for _ in range(N_REPEATS):
+        v = sample(n_leaves)
+
+        node1, node2 = np.random.choice(np.arange(2 * (n_leaves - 1)), 2, replace=False)
+
+        p2v_common_ancestor = get_common_ancestor(v, node1, node2)
+
+        nw = to_newick(v)
+
+        tr = Tree(nw, format=8)
+
+        ete3_common_ancestor = int(tr.get_common_ancestor(f"{node1}", f"{node2}").name)
+
+        assert p2v_common_ancestor == ete3_common_ancestor
+
+
 if __name__ == "__main__":
     pytest.main()

py-phylo2vec/phylo2vec/utils/__init__.py

@@ -11,7 +11,13 @@
 )
 from .random import sample, seed_everything
 from .validation import check_v
-from .vector import add_leaf, remove_leaf, reorder_v, reroot_at_random
+from .vector import (
+    add_leaf,
+    get_common_ancestor,
+    remove_leaf,
+    reorder_v,
+    reroot_at_random,
+)
 
 
 __all__ = [
@@ -20,6 +26,7 @@
     "check_v",
     "create_label_mapping",
     "find_num_leaves",
+    "get_common_ancestor",
     "remove_annotations",
     "remove_leaf",
     "remove_parent_labels",

py-phylo2vec/phylo2vec/utils/vector.py

@@ -254,6 +254,8 @@ def remove_leaf(v, leaf):
     -------
     v_sub : numpy.ndarray
         Phylo2Vec vector without `leaf`
+    sister : int
+        Sister node of leaf
     """
 
     # get the triplets from v
@@ -288,16 +290,19 @@ def remove_leaf(v, leaf):
 
     # We now have a correct ancestry without "leaf"
     # So we build a vector from it
-    cherries = _find_cherries(ancestry_sub)
 
     # Cherries have to be ordered according to the scheme presented in Fig. 2
+    # NOTE: not 100% sure why I need both orderings?
+    cherries = _find_cherries(ancestry_sub)
+    cherries_no_parents = _order_cherries_no_parents(cherries)
+
     # Build the new vector
-    v_sub = _build_vector(_order_cherries_no_parents(cherries))
+    v_sub = _build_vector(cherries_no_parents)
 
     return v_sub, sister
 
 
-@nb.njit
+@nb.njit(cache=True)
 def add_leaf(v, leaf, pos):
     """Add a leaf to a Phylo2Vec vector v
 
@@ -307,7 +312,7 @@ def add_leaf(v, leaf, pos):
         Phylo2Vec vector
     leaf : int >= 0
         A leaf node to add
-    leaf : int >= 0
+    pos : int >= 0
         A branch from where the leaf will be added
 
     Returns
@@ -340,9 +345,68 @@ def add_leaf(v, leaf, pos):
     ancestry_add[r_leaf, c_leaf] = leaf
 
     # Find the cherries
-    cherries = _order_cherries_no_parents(_find_cherries(ancestry_add))
+    # NOTE: not 100% sure why I need both orderings?
+    cherries = _find_cherries(ancestry_add)
+    cherries_no_parents = _order_cherries_no_parents(cherries)
 
     # Build the new vector
-    v_add = _build_vector(cherries)
+    v_add = _build_vector(cherries_no_parents)
 
     return v_add
+
+
+@nb.njit(cache=True)
+def get_ancestry_paths(v):
+    """
+    Get the ancestry paths for each node in the Phylo2Vec vector.
+
+    Parameters
+    ----------
+    v : numpy.ndarray
+        Phylo2Vec vector
+
+    Returns
+    -------
+    ancestry_paths : list of list of int
+        Ancestry paths for each node
+    """
+    ancestry = _get_ancestry(v)
+    parent_vec = np.zeros(2 * len(v), dtype=np.uint64)
+
+    for i in range(len(ancestry)):
+        parent_vec[ancestry[i, :2]] = ancestry[i, 2]
+
+    ancestry_paths = []
+    for i in range(2 * len(v)):
+        path = [i]
+        while (2 * len(v)) not in path:
+            path.insert(0, parent_vec[int(path[0])])
+        ancestry_paths.append(path)
+
+    return ancestry_paths
+
+
+@nb.njit(cache=True)
+def get_common_ancestor(v, node1, node2):
+    """Get the first recent common ancestor between two nodes in a Phylo2Vec tree
+
+    Parameters
+    ----------
+    v : numpy.ndarray
+        Phylo2Vec vector
+    node1 : int
+        A node in the tree
+    node2 : int
+        A node in the tree
+
+    Returns
+    -------
+    mrca : int
+        Most recent common ancestor node between node1 and node2
+    """
+    paths = get_ancestry_paths(v)
+    path1 = paths[node1]
+    path2 = paths[node2]
+    common_path = np.intersect1d(path1, path2)
+    mrca = common_path[0]
+    return mrca

py-phylo2vec/pyproject.toml

@@ -25,7 +25,7 @@ classifiers = [
 ]
 dependencies = [
   "numba>=0.56.4",
-  "numpy==1.24",
+  "numpy>=1.22,<2.1",
   "biopython==1.80.0",
   "joblib>=1.2.0",
   "ete3==3.1.3",

setup.py

@@ -11,7 +11,7 @@
 
 setup(
     name="phylo2vec",
-    version="0.1.10",
+    version="0.1.12",
     description="Phylo2Vec: integer vector representation of binary (phylogenetic) trees",
     long_description=LONG_DESCRIPTION,
     long_description_content_type="text/markdown",  # This is important!
@@ -22,7 +22,7 @@
     python_requires=">=3.9",
     install_requires=[
         "numba>=0.56.4",
-        "numpy==1.23.5",
+        "numpy>=1.22,<2.1",
         "biopython==1.80.0",
         "joblib>=1.2.0",
         "ete3==3.1.3",