Crashes sometimes; loads some values as strings

[iancrossfield]

When running Direct mode, I encountered this strange error -- it seems to be because values from the spreadsheet were loaded as strings ('0.039') instead of floats (0.039). Could be a problem with my CSV file, but it would be straightforward to just update ISOCLASSIFY to always cast the CSV-loaded values directly to type float.

Error trace:
.....
using max distance: 1026.2
using min distance: 907.1
Traceback (most recent call last):
File "/home/ianc/anaconda3/bin/isoclassify", line 8, in
sys.exit(main())
^^^^^^
File "/home/ianc/anaconda3/lib/python3.12/site-packages/isoclassify/isoclassify.py", line 67, in main
args.func(args)
File "/home/ianc/anaconda3/lib/python3.12/site-packages/isoclassify/isoclassify.py", line 70, in run
isoclassify.pipeline.run(**vars(args))
File "/home/ianc/anaconda3/lib/python3.12/site-packages/isoclassify/pipeline.py", line 36, in run
pipe.run()
File "/home/ianc/anaconda3/lib/python3.12/site-packages/isoclassify/pipeline.py", line 294, in run
self.paras = classify_direct.stparas(
^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ianc/anaconda3/lib/python3.12/site-packages/isoclassify/direct/classify.py", line 188, in stparas
map_samp = map + np.random.randn(nsample)*mape
~~~~~~~~~~~~~~~~~~~~~~~~^~~~~
numpy.core._exceptions._UFuncNoLoopError: ufunc 'multiply' did not contain a loop with signature matching types (dtype('float64'), dtype('<U5')) -> None