feat: add glass mock seed variants and v1.4.6 glass mock catalog

Add support for seed-specific mock catalog variants (e.g., SP_v1.4.5_glass_mock_seed1)
by extracting and substituting seed tokens in shear paths. Enables exploring multiple
random realizations of the same mock survey.

- Add SP_v1.4.6_glass_mock catalog entry with v1.4.6 survey specs
- Refactor version processing in __init__ to use recursive ensure_version_exists()
- Support _seed<N> variants that deep-copy base config and substitute seed token
- Handle _seed<N>_leak_corr combinations by materializing seed config first
- Add explicit error checking for missing seed tokens in paths
- Add regression tests for seed variant creation and error cases

Seed variant examples:
  - SP_v1.4.5_glass_mock_seed1 → unions_glass_sim_00001_4096.fits
  - SP_v1.4.6_glass_mock_seed12 → unions_glass_sim_00012_4096.fits
  - SP_v1.4.5_glass_mock_seed1_leak_corr → combines both transforms

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

Author: cailmdaley

notebooks/cosmo_val/cat_config.yaml

@@ -1070,6 +1070,50 @@ SP_v1.4.5_glass_mock:
     e2_col: e2
     path: unions_shapepipe_star_2024_v1.4.a.fits
 
+SP_v1.4.6_glass_mock:
+  subdir: /n17data/UNIONS/WL/v1.4.x
+  pipeline: SP
+  colour: darkgreen
+  getdist_colour: 0.0, 0.7, 0.0
+  ls: dashed
+  marker: d
+  cov_th:
+    A: 2405.3892055695346
+    n_e: 6.128201234871523
+    n_psf: 0.752316232272063
+    sigma_e: 0.379587601488189
+  mask: /home/guerrini/sp_validation/cosmo_inference/data/mask/mask_map_v1.4.6_nside_8192.fits
+  psf:
+    PSF_flag: FLAG_PSF_HSM
+    PSF_size: SIGMA_PSF_HSM
+    square_size: true
+    star_flag: FLAG_STAR_HSM
+    star_size: SIGMA_STAR_HSM
+    hdu: 1
+    path: unions_shapepipe_psf_2024_v1.4.a.fits
+    ra_col: RA
+    dec_col: Dec
+    e1_PSF_col: E1_PSF_HSM
+    e1_star_col: E1_STAR_HSM
+    e2_PSF_col: E2_PSF_HSM
+    e2_star_col: E2_STAR_HSM
+  shear:
+    R: 1.0
+    covmat_file: ./covs/shapepipe_A/cov_shapepipe_A.txt
+    path: /n09data/guerrini/glass_mock/results/unions_glass_sim_00000_4096.fits
+    redshift_path: /n17data/mkilbing/astro/data/CFIS/v1.0/nz/dndz_SP_A.txt
+    w_col: w
+    e1_col: e1
+    e1_PSF_col: e1_PSF
+    e2_col: e2
+    e2_PSF_col: e2_PSF
+  star:
+    ra_col: RA
+    dec_col: Dec
+    e1_col: e1
+    e2_col: e2
+    path: unions_shapepipe_star_2024_v1.4.a.fits
+
 SP_v1.4.5_intermediate:
   subdir: /n17data/murray/unions_cats
   pipeline: SP

src/sp_validation/cosmo_val.py

@@ -1,6 +1,7 @@
 # %%
 import copy
 import os
+import re
 from pathlib import Path
 
 import colorama
@@ -106,6 +107,34 @@ class CosmologyValidation:
     - TreeCorr cross_patch_weight: Automatically set to 'match' for jackknife,
       'simple' otherwise, following TreeCorr best practices.
     """
+
+    _SEED_SUFFIX_RE = re.compile(r"^(?P<base>.+)_seed(?P<seed>\d+)$")
+    _SEED_TOKEN_RE = re.compile(r"seed(?P<sep>[_-]?)(?P<digits>\d+)")
+
+    @classmethod
+    def _split_seed_variant(cls, version):
+        """Return the base version and seed label if version encodes a seed."""
+        match = cls._SEED_SUFFIX_RE.match(version)
+        if match is None:
+            return None, None
+        return match.group("base"), match.group("seed")
+
+    @classmethod
+    def _apply_seed_token(cls, path, seed_value, version, base_version, catalog_config):
+        """Replace the final seed token in the path with the requested seed value."""
+        matches = list(cls._SEED_TOKEN_RE.finditer(path))
+        if not matches:
+            raise ValueError(
+                f"Cannot materialize '{version}': shear path '{path}' for base version "
+                f"'{base_version}' does not contain a 'seed<digits>' token. "
+                f"Update {catalog_config} or drop the seed suffix."
+            )
+        match = matches[-1]
+        sep = match.group("sep") or ""
+        replacement = f"seed{sep}{seed_value}"
+        start, end = match.span()
+        return f"{path[:start]}{replacement}{path[end:]}"
+
     def __init__(
         self,
         versions,
@@ -191,38 +220,58 @@ def resolve_paths_for_version(ver):
         resolve_paths_for_version("nz")
         processed = {"nz"}
         final_versions = []
+        leak_suffix = "_leak_corr"
 
-        for ver in versions:
-            if ver.endswith("_leak_corr"):
-                base_ver = ver.replace("_leak_corr", "")
-                target = base_ver
-                if base_ver not in cc:
-                    raise KeyError(
-                        f"Base version {base_ver} not found for {ver} in config file "
-                        f"{catalog_config}"
-                    )
-                if "e1_col_corrected" not in cc[base_ver]["shear"]:
+        def ensure_version_exists(ver):
+            if ver in processed:
+                return
+
+            if ver in cc:
+                resolve_paths_for_version(ver)
+                processed.add(ver)
+                return
+
+            seed_base, seed_label = self._split_seed_variant(ver)
+
+            if ver.endswith(leak_suffix):
+                base_ver = ver[: -len(leak_suffix)]
+                ensure_version_exists(base_ver)
+                shear_cfg = cc[base_ver]["shear"]
+                if "e1_col_corrected" not in shear_cfg or "e2_col_corrected" not in shear_cfg:
                     raise ValueError(
                         f"{base_ver} does not have e1_col_corrected/e2_col_corrected "
                         f"fields; cannot create {ver}"
                     )
-            else:
-                target = ver
                 if ver not in cc:
-                    raise KeyError(
-                        f"Version string {ver} not found in config file "
-                        f"{catalog_config}"
+                    cc[ver] = copy.deepcopy(cc[base_ver])
+                    cc[ver]["shear"]["e1_col"] = shear_cfg["e1_col_corrected"]
+                    cc[ver]["shear"]["e2_col"] = shear_cfg["e2_col_corrected"]
+                resolve_paths_for_version(ver)
+                processed.add(ver)
+                return
+
+            if seed_base is not None:
+                ensure_version_exists(seed_base)
+                if ver not in cc:
+                    cc[ver] = copy.deepcopy(cc[seed_base])
+                    seed_path = self._apply_seed_token(
+                        cc[seed_base]["shear"]["path"],
+                        seed_label,
+                        ver,
+                        seed_base,
+                        catalog_config,
                     )
+                    cc[ver]["shear"]["path"] = seed_path
+                resolve_paths_for_version(ver)
+                processed.add(ver)
+                return
 
-            if target not in processed:
-                resolve_paths_for_version(target)
-                processed.add(target)
-
-            if ver.endswith("_leak_corr"):
-                cc[ver] = copy.deepcopy(cc[base_ver])
-                cc[ver]["shear"]["e1_col"] = cc[base_ver]["shear"]["e1_col_corrected"]
-                cc[ver]["shear"]["e2_col"] = cc[base_ver]["shear"]["e2_col_corrected"]
+            raise KeyError(
+                f"Version string {ver} not found in config file {catalog_config}"
+            )
 
+        for ver in versions:
+            ensure_version_exists(ver)
             final_versions.append(ver)
 
         self.versions = final_versions

src/sp_validation/tests/test_cosmo_val.py

@@ -11,6 +11,7 @@
 import os
 
 import pytest
+import yaml
 
 from sp_validation.cosmo_val import CosmologyValidation
 
@@ -42,6 +43,57 @@ def base_config(self, tmp_path):
             "nbins": 20,
         }
 
+    @staticmethod
+    def _make_seed_config(tmp_path, shear_filename):
+        """Create a minimal catalog config for seed variant testing."""
+        base_version = "TestCatalog"
+        base_dir = tmp_path / "catalog"
+        base_dir.mkdir()
+        (base_dir / shear_filename).touch()
+
+        star_filename = "star_seed_1234.fits"
+        (base_dir / star_filename).touch()
+
+        nz_dir = tmp_path / "nz"
+        nz_dir.mkdir()
+        (nz_dir / "dndz.txt").write_text("0.1 1.0\n")
+
+        output_dir = tmp_path / "output"
+        output_dir.mkdir()
+
+        config_path = tmp_path / "seed_config.yaml"
+        config_data = {
+            "nz": {
+                "subdir": str(nz_dir),
+                "dndz": {"blind": "A", "path": "dndz.txt"},
+            },
+            "paths": {"output": str(output_dir)},
+            base_version: {
+                "subdir": str(base_dir),
+                "pipeline": "SP",
+                "shear": {
+                    "path": shear_filename,
+                    "w_col": "w",
+                    "e1_col": "e1",
+                    "e2_col": "e2",
+                    "e1_col_corrected": "e1_corr",
+                    "e2_col_corrected": "e2_corr",
+                },
+                "star": {"path": star_filename},
+            },
+        }
+        config_path.write_text(yaml.dump(config_data, sort_keys=False))
+
+        params = {
+            "catalog_config": str(config_path),
+            "output_dir": str(output_dir),
+            "npatch": 1,
+            "theta_min": 1.0,
+            "theta_max": 250.0,
+            "nbins": 20,
+        }
+        return params, base_version
+
     @pytest.mark.parametrize(
         "version,e1_col,e2_col",
         [
@@ -125,3 +177,42 @@ def test_additive_bias_leak_corrected_columns(self, base_config, version):
         # Verify the values are numeric
         assert isinstance(cv.c1[version_leak_corr], float)
         assert isinstance(cv.c2[version_leak_corr], float)
+
+    def test_seed_variant_updates_shear_path(self, tmp_path):
+        """Seeded versions should materialize a seed-specific shear path."""
+        params, base_version = self._make_seed_config(
+            tmp_path, shear_filename="shear_seed_1234.fits"
+        )
+        seed_version = f"{base_version}_seed007"
+
+        cv = CosmologyValidation(versions=[seed_version], **params)
+
+        assert cv.versions == [seed_version]
+        assert seed_version in cv.cc
+        assert cv.cc[seed_version]["shear"]["path"].endswith("shear_seed_007.fits")
+
+    def test_seed_leak_corr_materializes_seed_first(self, tmp_path):
+        """_seed<N>_leak_corr should clone the seed variant before leak fixes."""
+        params, base_version = self._make_seed_config(
+            tmp_path, shear_filename="shear_seed_1234.fits"
+        )
+        leak_version = f"{base_version}_seed007_leak_corr"
+        seed_version = f"{base_version}_seed007"
+
+        cv = CosmologyValidation(versions=[leak_version], **params)
+
+        assert cv.versions == [leak_version]
+        assert seed_version in cv.cc
+        assert cv.cc[seed_version]["shear"]["path"].endswith("shear_seed_007.fits")
+        assert cv.cc[leak_version]["shear"]["e1_col"] == "e1_corr"
+        assert cv.cc[leak_version]["shear"]["e2_col"] == "e2_corr"
+
+    def test_seed_variant_without_token_errors(self, tmp_path):
+        """Missing seed token in shear path should raise a descriptive error."""
+        params, base_version = self._make_seed_config(
+            tmp_path, shear_filename="shear_base.fits"
+        )
+        seed_version = f"{base_version}_seed123"
+
+        with pytest.raises(ValueError, match="seed"):
+            CosmologyValidation(versions=[seed_version], **params)