Typing nit in asert_daa.py

Technically EC sometimes stores the ts as a float or an int (I'm unclear
on when it does either) -- so type annocation should reflect that
sometimes the ts can be int or float.  For the actual calculation it
doesn't matter so much.

lib/asert_daa.py

@@ -23,7 +23,7 @@
 import os
 
 from collections import namedtuple
-from typing import Optional
+from typing import Optional, Union
 
 from .util import print_error
 
@@ -111,7 +111,7 @@ def target_to_hex(target: int) -> str:
         h = hex(target)[2:]
         return '0' * (64 - len(h)) + h
 
-    def next_bits_aserti3_2d(self, anchor_bits: int, time_diff: float, height_diff: int) -> int:
+    def next_bits_aserti3_2d(self, anchor_bits: int, time_diff: Union[float, int], height_diff: int) -> int:
         """ Integer ASERTI algorithm, based on Jonathan Toomim's
         `next_bits_aserti` implementation in mining.py (see
         https://github.com/jtoomim/difficulty) """