🐛 binder notebook gdp-6hourly is broken
Closed this issue · 2 comments
malmans2 commented
Hi there,
I just tried the first binder notebook (gdp-6hourly.ipynb). Cell number 11 fails with this error:
---------------------------------------------------------------------------
KeyError Traceback (most recent call last)
File /srv/conda/envs/notebook/lib/python3.10/site-packages/awkward/_dispatch.py:60, in named_high_level_function.<locals>.dispatch(*args, **kwargs)
59 try:
---> 60 next(gen_or_result)
61 except StopIteration as err:
File /srv/conda/envs/notebook/lib/python3.10/site-packages/awkward/operations/ak_to_numpy.py:44, in to_numpy(array, allow_missing)
43 # Implementation
---> 44 return _impl(array, allow_missing)
File /srv/conda/envs/notebook/lib/python3.10/site-packages/awkward/operations/ak_to_numpy.py:56, in _impl(array, allow_missing)
54 numpy_layout = layout.to_backend(backend)
---> 56 return numpy_layout.to_backend_array(allow_missing=allow_missing)
File /srv/conda/envs/notebook/lib/python3.10/site-packages/awkward/contents/content.py:1077, in Content.to_backend_array(self, allow_missing, backend)
1076 backend = regularize_backend(backend)
-> 1077 return self._to_backend_array(allow_missing, backend)
File /srv/conda/envs/notebook/lib/python3.10/site-packages/awkward/contents/listoffsetarray.py:2042, in ListOffsetArray._to_backend_array(self, allow_missing, backend)
2040 buffer = backend.nplike.empty(max_count * self.length, dtype=np.uint8)
-> 2042 self.backend[
2043 "awkward_NumpyArray_pad_zero_to_length",
2044 self._content.dtype.type,
2045 self._offsets.dtype.type,
2046 buffer.dtype.type,
2047 ](
2048 self._content.data,
2049 self._offsets.data,
2050 self._offsets.length,
2051 max_count,
2052 buffer,
2053 )
2054 return buffer.view(np.dtype(("S", max_count)))
File /srv/conda/envs/notebook/lib/python3.10/site-packages/awkward/_backends/numpy.py:35, in NumpyBackend.__getitem__(self, index)
34 def __getitem__(self, index: KernelKeyType) -> NumpyKernel:
---> 35 return NumpyKernel(awkward_cpp.cpu_kernels.kernel[index], index)
KeyError: ('awkward_NumpyArray_pad_zero_to_length', <class 'numpy.uint8'>, <class 'numpy.int32'>, <class 'numpy.uint8'>)
The above exception was the direct cause of the following exception:
KeyError Traceback (most recent call last)
Cell In[11], line 1
----> 1 ra = RaggedArray.from_parquet("../data/process/gdp_6h.parquet")
File /srv/conda/envs/notebook/lib/python3.10/site-packages/clouddrift/raggedarray.py:166, in RaggedArray.from_parquet(cls, filename, name_coords)
148 @classmethod
149 def from_parquet(
150 cls, filename: str, name_coords: Optional[list] = ["time", "lon", "lat", "ids"]
151 ):
152 """Read a ragged array from a parquet file.
153
154 Parameters
(...)
164 A ragged array instance
165 """
--> 166 return cls.from_awkward(ak.from_parquet(filename), name_coords)
File /srv/conda/envs/notebook/lib/python3.10/site-packages/clouddrift/raggedarray.py:63, in RaggedArray.from_awkward(cls, array, name_coords)
60 attrs_variables[var] = array.obs[var].layout.parameters["attrs"]
62 for var in [v for v in array.fields if v != "obs"]:
---> 63 metadata[var] = array[var].to_numpy()
64 attrs_variables[var] = array[var].layout.parameters["attrs"]
66 for var in [v for v in array.obs.fields if v not in name_coords]:
File /srv/conda/envs/notebook/lib/python3.10/site-packages/awkward/highlevel.py:459, in Array.to_numpy(self, allow_missing)
455 def to_numpy(self, allow_missing=True):
456 """
457 Converts this Array into a NumPy array, if possible; same as #ak.to_numpy.
458 """
--> 459 return ak.operations.to_numpy(self, allow_missing=allow_missing)
File /srv/conda/envs/notebook/lib/python3.10/site-packages/awkward/_dispatch.py:36, in named_high_level_function.<locals>.dispatch(*args, **kwargs)
33 @wraps(func)
34 def dispatch(*args, **kwargs):
35 # NOTE: this decorator assumes that the operation is exposed under `ak.`
---> 36 with OperationErrorContext(name, args, kwargs):
37 gen_or_result = func(*args, **kwargs)
38 if isgenerator(gen_or_result):
File /srv/conda/envs/notebook/lib/python3.10/site-packages/awkward/_errors.py:67, in ErrorContext.__exit__(self, exception_type, exception_value, traceback)
60 try:
61 # Handle caught exception
62 if (
63 exception_type is not None
64 and issubclass(exception_type, Exception)
65 and self.primary() is self
66 ):
---> 67 self.handle_exception(exception_type, exception_value)
68 finally:
69 # `_kwargs` may hold cyclic references, that we really want to avoid
70 # as this can lead to large buffers remaining in memory for longer than absolutely necessary
71 # Let's just clear this, now.
72 self._kwargs.clear()
File /srv/conda/envs/notebook/lib/python3.10/site-packages/awkward/_errors.py:82, in ErrorContext.handle_exception(self, cls, exception)
80 self.decorate_exception(cls, exception)
81 else:
---> 82 raise self.decorate_exception(cls, exception)
KeyError: ('awkward_NumpyArray_pad_zero_to_length', <class 'numpy.uint8'>, <class 'numpy.int32'>, <class 'numpy.uint8'>)
This error occurred while calling
ak.to_numpy(
<Array [b'SVPB ', b'SVP ', ..., b'SVPB '] type='100 * bytes'>
allow_missing = True
)
I haven't tried running the notebook locally, but I noticed that an old version of clouddrift (0.21.2) is installed on Binder. This might be a similar issue to #449, where the usage code is outdated. If that's the case, I think you need to automate the way you handle and publish the demos.
Let me know when I can try again.
philippemiron commented
Once we release a new version, it should be easier to fix with #454.
selipot commented
We have re-evaluated the usefulness of this notebook and have decided that it was not the best to showcase the library and we have therefore made this repo s private for now with the intent on revisiting these later. The goals of these notebooks was to explain how to build data adapters for Lagrangian datasets but these methods are not destined for the use of new users of the library. In the readme of the library we now emphasize and link three other notebooks that are better for first users.