Update init matching procedure (#50)

* pep8

* fix convention

* Update script

* enforce optimisation boundaries to be finite

* Update TODO

* Remove sky matching

* FIx a small bug

* fix bug

* Remove import

* Add halo fitted quantities

* Update nbs

* update README

* Add load_initial comments

* Rename nbs

* Delete nb

* Update imports

* Rename function

* Update matcher

* Add overlap paths

* Update the matching script

* Update verbosity

* Add verbosity flags

* Simplify make_bckg_delta

* bug fix

* fix bug

* lala

* la

* Add overlap paths

* Update limit

* pep8

* Some pep8 stuff

* pep8

* Minor corrections

* Update paths

* Add number of particles check

* Fix bug

* pep8 comments

csiborgtools/match/match.py

@@ -115,9 +115,9 @@ class RealisationsMatcher:
 
         Parameters
         ----------
-        cat0 : :py:class:`csiborgtools.read.ClumpsCatalogue`
+        cat0 : :py:class:`csiborgtools.read.HaloCatalogue`
             Halo catalogue of the reference simulation.
-        catx : :py:class:`csiborgtools.read.ClumpsCatalogue`
+        catx : :py:class:`csiborgtools.read.HaloCatalogue`
             Halo catalogue of the cross simulation.
         halos0_archive : `NpzFile` object
             Archive of halos' particles of the reference simulation, keys must
@@ -198,10 +198,8 @@ class RealisationsMatcher:
                 except KeyError:
                     halox = halosx_archive[str(kf)]
                     minsx, maxsx = get_halolims(
-                        halox,
-                        ncells=self.overlapper.inv_clength,
-                        nshift=self.overlapper.nshift,
-                    )
+                        halox, ncells=self.overlapper.inv_clength,
+                        nshift=self.overlapper.nshift)
                     for p in ("x", "y", "z"):
                         halox[p] = self.overlapper.pos2cell(halox[p])
                     cross_halos[kf] = halox
@@ -768,9 +766,8 @@ def calculate_overlap(delta1, delta2, cellmins, delta_bckg):
 
 
 @jit(nopython=True)
-def calculate_overlap_indxs(
-    delta1, delta2, cellmins, delta_bckg, nonzero, mass1, mass2
-):
+def calculate_overlap_indxs(delta1, delta2, cellmins, delta_bckg, nonzero,
+                            mass1, mass2):
     r"""
     Overlap between two clumps whose density fields are evaluated on the
     same grid and `nonzero1` enumerates the non-zero cells of `delta1.  This is
@@ -841,9 +838,8 @@ def dist_centmass(clump):
         Center of mass coordinates.
     """
     # CM along each dimension
-    cmx, cmy, cmz = [
-        numpy.average(clump[p], weights=clump["M"]) for p in ("x", "y", "z")
-    ]
+    cmx, cmy, cmz = [numpy.average(clump[p], weights=clump["M"])
+                     for p in ("x", "y", "z")]
     # Particle distance from the CM
     dist = numpy.sqrt(
         numpy.square(clump["x"] - cmx)
@@ -876,9 +872,8 @@ def dist_percentile(dist, qs, distmax=0.075):
     return x
 
 
-def radius_neighbours(
-    knn, X, radiusX, radiusKNN, nmult=1.0, enforce_int32=False, verbose=True
-):
+def radius_neighbours(knn, X, radiusX, radiusKNN, nmult=1.0,
+                      enforce_int32=False, verbose=True):
     """
     Find all neigbours of a trained KNN model whose center of mass separation
     is less than `nmult` times the sum of their respective radii.

csiborgtools/read/overlap_summary.py

@@ -27,12 +27,12 @@ class PairOverlap:
 
     Parameters
     ----------
-    cat0, catx: :py:class:`csiborgtools.read.ClumpsCatalogue`
-        Halo catalogues corresponding to the reference and cross
-        simulations.
-    fskel : str, optional
-        Path to the overlap. By default `None`, i.e.
-        `/mnt/extraspace/rstiskalek/csiborg/overlap/cross_{}_{}.npz`.
+    cat0 : :py:class:`csiborgtools.read.HaloCatalogue`
+        Halo catalogue corresponding to the reference simulation.
+    catx : :py:class:`csiborgtools.read.HaloCatalogue`
+        Halo catalogue corresponding to the cross simulation.
+    paths : py:class`csiborgtools.read.CSiBORGPaths`
+        CSiBORG paths object.
     min_mass : float, optional
         Minimum :math:`M_{\rm tot} / M_\odot` mass in the reference catalogue.
         By default no threshold.
@@ -44,7 +44,7 @@ class PairOverlap:
     _catx = None
     _data = None
 
-    def __init__(self, cat0, catx, fskel=None, min_mass=None, max_dist=None):
+    def __init__(self, cat0, catx, paths, min_mass=None, max_dist=None):
         self._cat0 = cat0
         self._catx = catx
 

scripts/pre_initmatch.py

@@ -93,7 +93,7 @@ for i, nsim in enumerate(paths.get_ics(tonew=True)):
     # the end save these.
     cat = csiborgtools.read.ClumpsCatalogue(nsim, paths, load_fitted=False,
                                             rawdata=True)
-    parent_ids = cat["index"][cat.ismain][:500]
+    parent_ids = cat["index"][cat.ismain]
     jobs = csiborgtools.fits.split_jobs(parent_ids.size, nproc)[rank]
     for i in tqdm(jobs) if verbose else jobs:
         clid = parent_ids[i]
@@ -101,6 +101,9 @@ for i, nsim in enumerate(paths.get_ics(tonew=True)):
 
         mmain_mask = numpy.isin(clump_ids, mmain_indxs, assume_unique=True)
         mmain_particles = part0[mmain_mask]
+        # If the number of particles is too small, we skip this halo.
+        if mmain_particles.size < 100:
+            continue
 
         raddist, cmpos = csiborgtools.match.dist_centmass(mmain_particles)
         patchsize = csiborgtools.match.dist_percentile(raddist, [99],