Merge pull request #564 from jeromekelleher/yet-more-packaging

Add workflow for testing base API

Author: jeromekelleher

.github/workflows/ci.yml

@@ -86,6 +86,25 @@ jobs:
         run: |
           sc2ts info-matches testrun/test.matches.db
 
+  bare_api:
+    name: Bare API tests
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: '3.11'
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          python -m pip install '.'
+          python -m pip install pytest
+      - name: Run tests
+        run: |
+          # We are careful to pull in the minimum dependencies here
+          # and only operate on files stored in the repo
+          python -m pytest -v -c /dev/null -p no:sc2ts_fixtures tests/test_api.py
+
   packaging:
     name: Packaging tests
     runs-on: ubuntu-latest

sc2ts/jit.py

@@ -1,7 +1,6 @@
 import dataclasses
 import logging
 
-
 import numpy as np
 import numba
 import tskit.jit.numba as tskit_numba
@@ -94,8 +93,6 @@ def get_path_mrca(path1, path2, node_time):
     )
 
 
-
-
 @dataclasses.dataclass
 class ArgCounts:
     nodes_max_descendant_samples: np.ndarray
@@ -127,7 +124,9 @@ def _compute_inheritance_counts(
     edges_child = numba_ts.edges_child
     mutations_node = numba_ts.mutations_node
     mutations_parent = numba_ts.mutations_parent
-    mutations_position = numba_ts.sites_position[numba_ts.mutations_site].astype(np.int32)
+    mutations_position = numba_ts.sites_position[numba_ts.mutations_site].astype(
+        np.int32
+    )
     nodes_flags = numba_ts.nodes_flags
 
     parent = np.zeros(num_nodes, dtype=np.int32) - 1
@@ -202,7 +201,7 @@ def count(ts):
     )
 
 
-# FIXME make cache optional
+# FIXME make cache optional.
 @numba.njit(cache=True)
 def encode_alignment(h):
     # Just so numba knows this is a constant string.
@@ -220,4 +219,3 @@ def encode_alignment(h):
                 raise ValueError(f"Allele {h[j]} not recognised")
             a[j] = k
     return a
-

sc2ts/stats.py

@@ -1,5 +1,4 @@
 from . import core
-from . import jit
 
 import numpy as np
 import pandas as pd
@@ -15,7 +14,14 @@ def convert_date(ts, time_array):
     return time_zero_date - time_array.astype("timedelta64[D]")
 
 
-def node_data(ts, inheritance_stats=True):
+def run_count(ts):
+    # TODO Raise a friendly numba-error here
+    from . import jit
+
+    return jit.count(ts)
+
+
+def node_data(ts, inheritance_stats=False):
     """
     Return a pandas dataframe with one row for each node (in node ID order)
     from the specified tree sequence. This must be the output of
@@ -41,7 +47,7 @@ def node_data(ts, inheritance_stats=True):
     #         minlength=ts.num_edges)
 
     if inheritance_stats:
-        counter = jit.count(ts)
+        counter = run_count(ts)
         cols["max_descendant_samples"] = counter.nodes_max_descendant_samples
         dtype["max_descendant_samples"] = "int"
     if "time_zero_date" in ts.metadata:
@@ -50,7 +56,7 @@ def node_data(ts, inheritance_stats=True):
     return pd.DataFrame(cols).astype(dtype)
 
 
-def mutation_data(ts, inheritance_stats=True, parsimony_stats=False):
+def mutation_data(ts, inheritance_stats=False, parsimony_stats=False):
     """
     Return a pandas dataframe with one row for each mutation (in mutation ID order)
     from the specified tree sequence. This must be the output of
@@ -70,7 +76,7 @@ def mutation_data(ts, inheritance_stats=True, parsimony_stats=False):
     if not isinstance(ts.metadata, bytes) and "time_zero_date" in ts.metadata:
         cols["date"] = convert_date(ts, ts.mutations_time)
     if inheritance_stats:
-        counter = jit.count(ts)
+        counter = run_count(ts)
         cols["num_descendants"] = counter.mutations_num_descendants
         cols["num_inheritors"] = counter.mutations_num_inheritors