Faulty assignation of samples

[llrs]

There were two samples that the allocation was not correct.
This is related to the data from #47 (Olink-NeuroEpiCas).
In each plate (index 1) there were two samples to the same well.

It could be either spatial is faulty or inspect is faulty.
I could manually resolve the issue by using position_name.
The reason seems that plate 1 didn't use the well that was assigned to plate 2.
And vice versa, so both plates ended up with problems.

Consider exporting position_name and add more internal checks of the output of spatial and inspect.

[llrs]

Also check and reorder the spatial index according to positions

[llrs]

position_name returns a factor with all the possible positions so that will help to sort them.
The commit fixed the issue with the code and it is simpler and faster.