Improve and measure speed of clamp function

Change the comment at the start of the clamp
function because the comparison between the clip
function and the alternatives has been made.
Substitute the clamp function with the most
performant alternative.

mesa/space.py

@@ -56,9 +56,8 @@
 
 
 def clamp(x: float, lowest: float, highest: float) -> float:
-    # This should be faster than np.clip for a scalar x.
-    # TODO: measure how much faster this function is.
-    return max(lowest, min(x, highest))
+    # much faster than np.clip for a scalar x.
+    return lowest if x <= lowest else (highest if x >= highest else x)
 
 
 def accept_tuple_argument(wrapped_function: F) -> F: