update fam mask generation

src/fraposa_pgsc/fraposa.py

@@ -147,7 +147,8 @@ def read_bed(bed_filepref, dtype=np.int8, filt_iid=None):
             raise ValueError("Samples with duplicated FID + IID detected, please remove and retry")
 
         bed = np.zeros(shape=(p, len(matched_ids)), dtype=dtype)
-        fam_mask = fam.iid.isin((x.IID for x in matched_ids)) & fam.fid.isin(x.FID for x in matched_ids)
+        # in will call SampleID's __hash__ method which uses (fid, iid)
+        fam_mask = pd.Series((x in fam_ids for x in matched_ids))
         i_extract = np.where(fam_mask == True)
         for (i, (snp, genotypes)) in enumerate(pyp):
             bed[i,:] = genotypes[i_extract]

src/fraposa_pgsc/sampleid.py

@@ -1,8 +1,28 @@
-from dataclasses import dataclass
+class SampleID:
+    """ A sample ID from a plink fam file, including FID and IID """
+    def __init__(self, fid, iid):
+        self._fid = fid
+        self._iid = iid
 
+    def __repr__(self):
+        return f"{self.__class__.__name__}(fid={repr(self.fid)}, iid={repr(self.iid)})"
 
-@dataclass(frozen=True)
-class SampleID:
-    """ A sample ID from a plink fam/psam file, including FID and IID """
-    FID: str
-    IID: str
+    @property
+    def fid(self):
+        if self._fid == "0":  # 0 means missing :)
+            return self._iid
+        else:
+            return self._fid
+
+    @property
+    def iid(self):
+        return self._iid
+
+    def __hash__(self):
+        return hash((self.fid, self.iid))
+
+    def __eq__(self, other):
+        if not isinstance(other, SampleID):
+            return NotImplemented
+
+        return self.fid == other.fid and self.iid == other.iid