issue_comments: 112984626

This data as json

html_url issue_url id node_id user created_at updated_at author_association body reactions performed_via_github_app issue
https://github.com/pydata/xarray/issues/435#issuecomment-112984626 https://api.github.com/repos/pydata/xarray/issues/435 112984626 MDEyOklzc3VlQ29tbWVudDExMjk4NDYyNg== 6405510 2015-06-18T00:16:19Z 2015-06-18T00:16:19Z NONE

xray definitely seems to be the correct tool, as you suggested.

For the record, this is my first pass at coming up with the Dataset:

<xray.Dataset> Dimensions: (P: 20, T: 20, components: 4, id: 600, internal_dof: 9) Coordinates: * components (components) <U2 'AL' 'NI' 'CR' 'FE' * internal_dof (internal_dof) <U4 'AL_0' 'NI_0' 'CR_0' 'FE_0' 'AL_1' ... composition (id, components) float64 0.153 0.2138 0.2917 0.3415 0.316 ... Phase <U6 'FCC_A1' * P (P) float64 1e+05 1.833e+05 3.36e+05 6.158e+05 1.129e+06 ... * T (T) float64 300.0 347.4 394.7 442.1 489.5 536.8 584.2 ... * id (id) int64 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 ... Data variables: constitution (id, internal_dof) float64 0.1296 0.2055 0.301 0.3639 ... energies (T, P, id) float64 -5.533e+03 -605.8 -2.507e+03 -8.546e+03 ...

Full notebook: https://github.com/richardotis/pycalphad/blob/178f150b492099c32e197b417c11729f12d6dfe8/research/xrayTest.ipynb

I decided I'm better off giving each phase its own Dataset, and when I need to do multi-phase operations I'll drop all the internal dimensions before I merge them. The result of to_dataframe(), with how it's getting rows and columns mixed up, makes me think I don't yet specify the optimal combination of coordinates and variables in the Dataset.

Some initial queries of the data seem to function well and at a fraction of the memory cost of the pandas-based approach, so I'm feeling optimistic here.

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   88868867