Closed olamb245 closed 1 year ago
I just encountered this as well. We'll make sure it is fixed in an upcoming release. Assuming you worked around this already but you can just modify that line to be if n <= 2*p and alpha<1.0:
Yes, I think I came up with a fix very similar to this. Thanks.
Fixed by #23
I get the following error message when I use data from a 3-element array:
Bad assumption, n ! > 2p
The error message seems to come from line 79 in 'fast_lts_array'. I'm currently running 'ltsva' with Alpha=1, so it's running with ordinary least squares, not least-trimmed squares. I would understand needing the above error message if you're running least-trimmed with only 3 stations, but I'm not sure why it's needed for ordinary least squares when 3 stations should theoretically work (as far as I understand)?