Skipspecies

bin/create_csv_vectors_file.py

@@ -4,9 +4,9 @@
 import argparse
 from sys import argv
 
-from pymatgen.core import Element
+from pymatgen.core import Element, Species
 
-from skipatom import SkipAtomInducedModel, SkipAtomModel
+from skipatom import SkipAtomInducedModel, SkipAtomModel, SkipSpeciesInducedModel
 
 """
 e.g. 
@@ -43,6 +43,9 @@
     parser.add_argument(
         "--induced", action="store_true", help="whether to use induced SkipAtom vectors"
     )
+    parser.add_argument(
+        "--skipspecies", action="store_true", help="whether to use SkipSpecies vectors"
+    )
     parser.add_argument(
         "--min-count",
         required=("induced" in argv),
@@ -59,24 +62,45 @@
     args = parser.parse_args()
 
     if args.induced:
-        model = SkipAtomInducedModel.load(
-            args.model, args.data, min_count=args.min_count, top_n=args.top_n
-        )
+        if args.skipspecies:
+            model = SkipSpeciesInducedModel.load(
+                args.model, args.data, min_count=args.min_count, top_n=args.top_n
+            )
+        else:
+            model = SkipAtomInducedModel.load(
+                args.model, args.data, min_count=args.min_count, top_n=args.top_n
+            )
     else:
         model = SkipAtomModel.load(args.model, args.data)
 
-    sorted_elems = sorted(
-        [(e, Element(e).number) for e in model.dictionary], key=lambda v: v[1]
-    )
+    if args.skipspecies:
+        sorted_specs = sorted(
+            [(s, Species.from_string(s).number) for s in model.dictionary],
+            key=lambda v: v[1],
+        )
+    else:
+        sorted_elems = sorted(
+            [(e, Element(e).number) for e in model.dictionary], key=lambda v: v[1]
+        )
 
     dim = len(model.vectors[0])
 
     with open(args.out, "w") as f:
-        header = ["element"]
-        header.extend([str(i) for i in range(dim)])
-        f.write("%s\n" % ",".join(header))
-        for elem, _ in sorted_elems:
-            vec = model.vectors[model.dictionary[elem]].tolist()
-            row = [elem]
-            row.extend([str(v) for v in vec])
-            f.write("%s\n" % ",".join(row))
+        if args.skipspecies:
+            header = ["species"]
+            header.extend([str(i) for i in range(dim)])
+            f.write("%s\n" % ",".join(header))
+            for spec, _ in sorted_specs:
+                vec = model.vectors[model.dictionary[spec]].tolist()
+                row = [spec]
+                row.extend([str(v) for v in vec])
+                f.write("%s\n" % ",".join(row))
+        else:
+            header = ["element"]
+            header.extend([str(i) for i in range(dim)])
+            f.write("%s\n" % ",".join(header))
+            for elem, _ in sorted_elems:
+                vec = model.vectors[model.dictionary[elem]].tolist()
+                row = [elem]
+                row.extend([str(v) for v in vec])
+                f.write("%s\n" % ",".join(row))