phylogenie 1.0.6__tar.gz → 1.0.8__tar.gz

{phylogenie-1.0.6 → phylogenie-1.0.8}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: phylogenie
-Version: 1.0.6
+Version: 1.0.8
 Summary: Generate phylogenetic datasets with minimal setup effort
 Author: Gabriele Marino
 Author-email: gabmarino.8601@gmail.com

{phylogenie-1.0.6 → phylogenie-1.0.8}/phylogenie/backend/remaster/generate.py

@@ -24,7 +24,7 @@ def _beautify_xml(xml: bytes) -> str:
 
 def _generate_config_file(
     output_xml_file: str,
-    tree_file_name: str,
+    tree_filename: str,
     populations: list[str],
     init_values: list[int],
     sample_population: str,
@@ -85,7 +85,7 @@ def _generate_config_file(
     simulate.append(trajectory)
 
     logger = Element(
-        "logger", {"spec": "Logger", "mode": "tree", "fileName": tree_file_name}
+        "logger", {"spec": "Logger", "mode": "tree", "fileName": tree_filename}
     )
     logger.append(
         Element(
@@ -140,7 +140,7 @@ def _postprocess_tree(input_file: str, output_file: str, attributes: list[str])
 
 
 def generate_trees(
-    tree_file_name: str,
+    tree_filename: str,
     populations: str | list[str] = DEFAULT_POPULATION,
     init_population: str = DEFAULT_POPULATION,
     sample_population: str = SAMPLE_POPULATION,
@@ -158,14 +158,14 @@ def generate_trees(
     init_values[populations.index(init_population)] = 1
 
     if output_xml_file is None:
-        xml_file = f"{tree_file_name}-temp.xml"
+        xml_file = f"{tree_filename}-temp.xml"
     else:
         xml_file = output_xml_file
 
-    temp_tree_file = f"{tree_file_name}-temp.nex"
+    temp_tree_filename = f"{tree_filename}-temp.nex"
     _generate_config_file(
         output_xml_file=xml_file,
-        tree_file_name=temp_tree_file,
+        tree_filename=temp_tree_filename,
         populations=populations,
         init_values=init_values,
         sample_population=sample_population,
@@ -181,7 +181,7 @@ def generate_trees(
     cmd.append(xml_file)
     subprocess.run(cmd, check=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
 
-    _postprocess_tree(temp_tree_file, tree_file_name, ["type", "time"])
+    _postprocess_tree(temp_tree_filename, tree_filename, ["type", "time"])
     if output_xml_file is None:
         subprocess.run(["rm", xml_file], check=True)
-    subprocess.run(["rm", temp_tree_file], check=True)
+    subprocess.run(["rm", temp_tree_filename], check=True)

{phylogenie-1.0.6 → phylogenie-1.0.8}/phylogenie/backend/treesimulator.py

@@ -101,6 +101,7 @@ def generate_tree(
     )
     for i, leaf in enumerate(tree.iter_leaves()):
         state: str = getattr(leaf, STATE)
+        delattr(leaf, STATE)
         date = tree.get_distance(leaf)
         leaf.name = f"{i}|{state}|{date}"
     save_forest([tree], output_file)

{phylogenie-1.0.6 → phylogenie-1.0.8}/phylogenie/core/msas/base.py

@@ -1,6 +1,7 @@
 import os
 from abc import abstractmethod
 from enum import Enum
+from pathlib import Path
 from typing import Literal
 
 from numpy.random import Generator
@@ -14,9 +15,14 @@ class BackendType(str, Enum):
     ALISIM = "alisim"
 
 
+MSAS_DIRNAME = "MSAs"
+TREES_DIRNAME = "trees"
+
+
 class MSAsGenerator(DatasetGenerator):
     data_type: Literal[DataType.MSAS] = DataType.MSAS
     trees: TreesGeneratorConfig
+    keep_trees: bool = False
 
     @abstractmethod
     def _generate_one_from_tree(
@@ -24,11 +30,22 @@ class MSAsGenerator(DatasetGenerator):
     ) -> None: ...
 
     def _generate_one(self, filename: str, rng: Generator, data: pgt.Data) -> None:
-        tree_filename = f"{filename}.temp-tree"
+        if self.keep_trees:
+            base_dir = Path(filename).parent
+            file_id = Path(filename).stem
+            tree_filename = os.path.join(base_dir, TREES_DIRNAME, file_id)
+            msas_dir = os.path.join(base_dir, MSAS_DIRNAME)
+            os.makedirs(msas_dir, exist_ok=True)
+            msa_filename = os.path.join(msas_dir, file_id)
+        else:
+            tree_filename = f"{filename}.temp-tree"
+            msa_filename = filename
+
         self.trees.generate_one(
             filename=tree_filename, data=data, seed=int(rng.integers(0, 2**32 - 1))
         )
         self._generate_one_from_tree(
-            filename=filename, tree_file=f"{tree_filename}.nwk", rng=rng, data=data
+            filename=msa_filename, tree_file=f"{tree_filename}.nwk", rng=rng, data=data
         )
-        os.remove(f"{tree_filename}.nwk")
+        if not self.keep_trees:
+            os.remove(f"{tree_filename}.nwk")

{phylogenie-1.0.6 → phylogenie-1.0.8}/phylogenie/core/trees/remaster/generator.py

@@ -51,7 +51,7 @@ class ReMASTERGenerator(TreesGenerator):
         self, filename: str, rng: Generator, data: pgt.Data, reactions: list[Reaction]
     ) -> None:
         generate_trees(
-            tree_file_name=f"{filename}.nwk",
+            tree_filename=f"{filename}.nwk",
             populations=self.populations,
             init_population=self.init_population.format(**data),
             sample_population=self.sample_population,

{phylogenie-1.0.6 → phylogenie-1.0.8}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "phylogenie"
-version = "1.0.6"
+version = "1.0.8"
 description = "Generate phylogenetic datasets with minimal setup effort"
 authors = ["Gabriele Marino <gabmarino.8601@gmail.com>"]
 readme = "README.md"