I may have found a non -tame example in a bounded space.

Let our data set be . Construct the cover by where

and . Notice that and . Moreover, under the Lebesgue measure ,

Thus, in the Steinhaus filtration the edge is born at time

So all the edges are born at the same time, which is not what we want.

Either way, the first-homology group of this complex is the trivial group as there are no 1-cycles.

However, we might be able to take this idea to the non -tame example from 2024-10-10. The idea is to make each diamond have a height of .

We remove the bottom half of the diamonds and compress the remaining top half into a box. The boxes should have the following sizes:

The overlap , , and should be as follows:

Assuming this is even possible, we get the Steinhaus distances of

and

Therefore, in the first-homology persistence module the map from to will have infinite rank. Unsure if this is true.