Infinite work if in dataset presents a point with a sharp deviation in all dimensions

[valera7979]

I tried to test application on 10 dimensional points. But did not wait for the result. Dataset was very simple - only 3 points.
1;1;1;1;1;1;1;1;1;1
2;2;2;2;2;2;2;1;1;1
11;11;11;22;22;22;33;22;22;22

I found condition when it works. If value in one dimension is close to all other points - it works.

1;1;1;1;1;1;1;1;1;1
2;2;2;2;2;2;2;1;1;1
11;11;11;22;22;22;33;22;22;2 <--

And i found also, the more dimnsions the less deviation is required for such behavior.

[valera7979]

I found little workaround. When added one column with 0 in the middle of the table it works. But i don`t know why))