Issue in no data zone in line.py line 196

[bullocke]

https://github.com/ceholden/yatsm/blob/master/yatsm/cli/line.py#L196

File "/usr3/graduate/valpasq/Documents/yatsm/yatsm/cli/line.py", line 196, in line
output.extend(yatsm.record)
TypeError: 'NoneType' object is not iterable

Fix (tested):
Add if statement to check if record exists:

if yatsm.record is not None:
output.extend(yatsm.record)

[ceholden]

Doesn't this bit of code address the issue? It does effectively the same thing but it is placed before the phenology calculation and refitting?

https://github.com/ceholden/yatsm/blob/master/yatsm/cli/line.py#L176

if yatsm.record is None:
    continue
...
output.extend(yatsm.record)

It seems like something else is going on since it should skip to the next column if the yatsm.record is None.

[ceholden]

Oops -- so this is basically a copy of #47. The fix I implemented didn't actually work:

When I check, if yatsm.record is None, yatsm.record isn't actually None -- it's just an empty np.array. The refit or postprocess functions actually return None if the yatsm.record is empty so this is why your suggestion would work. I'm going to fix it even earlier on by actually correctly checking for an empty result.

[bullocke]

Makes sense, please notify when ready for checking and we will test.

[ceholden]

I tried it on Val's dataset and it worked. Can you give it a go? If it's working I'll re-tag this commit as v0.5.3 and re-release it.

[valpasq]

I pulled the latest fix, currently running on the /Massachusetts/p012r031/ dataset without issue, i.e. no error messages, YATSM results are being written. I think this issue can be closed.

[ceholden]

Sorry about that! Anyway as I mentioned I'm going to re-release v0.5.3 by moving it one step further to this commit.