Relaxed the requirements for calling find_min_box_constrained() and

find_max_box_constrained().  Now the bounds can be empty for some variables.

dlib/optimization/optimization.h

@@ -482,7 +482,7 @@ namespace dlib
             << "\n\t x_upper.size():         " << x_upper.size()
         );
         DLIB_ASSERT (
-            min(x_upper-x_lower) > 0,
+            min(x_upper-x_lower) >= 0,
             "\tdouble find_min_box_constrained()"
             << "\n\t You have to supply proper box constraints to this function."
             << "\n\r min(x_upper-x_lower): " << min(x_upper-x_lower)
@@ -610,7 +610,7 @@ namespace dlib
             << "\n\t x_upper.size():         " << x_upper.size()
         );
         DLIB_ASSERT (
-            min(x_upper-x_lower) > 0,
+            min(x_upper-x_lower) >= 0,
             "\tdouble find_max_box_constrained()"
             << "\n\t You have to supply proper box constraints to this function."
             << "\n\r min(x_upper-x_lower): " << min(x_upper-x_lower)

dlib/optimization/optimization_abstract.h

@@ -297,7 +297,7 @@ namespace dlib
             - is_col_vector(x_upper) == true
             - x.size() == x_lower.size() == x_upper.size()
               (i.e. x, x_lower, and x_upper need to all be column vectors of the same dimensionality)
-            - min(x_upper-x_lower) > 0
+            - min(x_upper-x_lower) >= 0
               (i.e. x_upper must contain upper bounds relative to x_lower)
         ensures
             - Performs a box constrained minimization of the function f() using the given
@@ -391,7 +391,7 @@ namespace dlib
             - is_col_vector(x_upper) == true
             - x.size() == x_lower.size() == x_upper.size()
               (i.e. x, x_lower, and x_upper need to all be column vectors of the same dimensionality)
-            - min(x_upper-x_lower) > 0
+            - min(x_upper-x_lower) >= 0
               (i.e. x_upper must contain upper bounds relative to x_lower)
         ensures
             - Performs a box constrained maximization of the function f() using the given