search

Quansight / lsst_dashboard

LSST Dashboard https://quansight.github.io/lsst_dashboard/
BSD 3-Clause "New" or "Revised" License
8 stars 3 forks source link

infs are still in data #117

Closed timothydmorton closed 4 years ago

timothydmorton commented 4 years ago

Testing out the dashboard on the new stack on lsst-dev, I get the following error:

te-packages/holoviews/core/operation.py", line 143, in process_element                             |
    return self._apply(element, key)                                                               |
  File "/software/lsstsw/stack_20200220/python/miniconda3-4.7.12/envs/lsst-scipipe/lib/python3.7/si|
te-packages/holoviews/core/operation.py", line 121, in _apply                                      |
    ret = self._process(element, key)                                                              |
  File "/software/lsstsw/stack_20200220/python/miniconda3-4.7.12/envs/lsst-scipipe/lib/python3.7/si|
te-packages/holoviews/operation/datashader.py", line 1115, in _process                             |
    img = tf.shade(array, **shade_opts)                                                            |
  File "/software/lsstsw/stack_20200220/python/miniconda3-4.7.12/envs/lsst-scipipe/lib/python3.7/si|
te-packages/datashader/transfer_functions.py", line 370, in shade                                  |
    return _interpolate(agg, cmap, how, alpha, span, min_alpha, name)                              |
  File "/software/lsstsw/stack_20200220/python/miniconda3-4.7.12/envs/lsst-scipipe/lib/python3.7/si|
te-packages/datashader/transfer_functions.py", line 215, in _interpolate                           |
    data = interpolater(interp, mask)                                                              |
  File "/software/lsstsw/stack_20200220/python/miniconda3-4.7.12/envs/lsst-scipipe/lib/python3.7/si|
te-packages/datashader/transfer_functions.py", line 155, in eq_hist                                |
    hist, bin_edges = np.histogram(data2, bins=nbins)                                              |
  File "<__array_function__ internals>", line 6, in histogram                                      |
  File "/software/lsstsw/stack_20200220/python/miniconda3-4.7.12/envs/lsst-scipipe/lib/python3.7/si|
te-packages/numpy/lib/histograms.py", line 795, in histogram                                       |
    bin_edges, uniform_bins = _get_bin_edges(a, bins, range, weights)                              |
  File "/software/lsstsw/stack_20200220/python/miniconda3-4.7.12/envs/lsst-scipipe/lib/python3.7/si|
te-packages/numpy/lib/histograms.py", line 429, in _get_bin_edges                                  |
    first_edge, last_edge = _get_outer_edges(a, range)                                             |
  File "/software/lsstsw/stack_20200220/python/miniconda3-4.7.12/envs/lsst-scipipe/lib/python3.7/si|
te-packages/numpy/lib/histograms.py", line 327, in _get_outer_edges                                |
    "autodetected range of [{}, {}] is not finite".format(first_edge, last_edge))                  |
ValueError: autodetected range of [0.0, inf] is not finite

There shouldn't be any infs left in these datasets...

timothydmorton commented 4 years ago

I confirm same issue exists on my local machine using current environment.yml

timothydmorton commented 4 years ago

This needs to pass df, not self.df: https://github.com/Quansight/lsst_dashboard/blob/2030010a635ef80ca4cf624f6f7a485a4db6d32b/lsst_dashboard/qa_dataset.py#L129