heliopython / heliopy Goto Github PK

See ftp://apollo.ssl.berkeley.edu/pub/helios-data/

xref conda-forge/staged-recipes#4292

This should speed up loading data.

Currently HelioPy uses a copy of the cdf reader from SpacePy. This isn't ideal because we don't want to maintain our own cdf reader, and the reader should really exist in an independent package.

It currently errors on large intervals due to recursion

A neat little package to automatically summarise modules: https://github.com/astropy/sphinx-automodapi

• FGM magnetic field
• PEACE particle moments
• STAFF magnetic field
• EFW electric field
• CIS particle moments

Currently it's parsed as a string, but it should be a bool

This was taken out as part of #58, but could be put back in as an option.

Add separate methods for protons, electrons, and the distribution files at the top of the ascii files. Should speed up loading if only 1/2 of these is wanted.

There's a few locations that it happens whatever the use_hdf config variable is, here's a reminder to fix those.

Currently there isn't much documentation about the new spice module, some more should be added.

eg. electron pitch angle distributions imported from WIND.

... and also no extra time / ordinal columns

Currently rely on spacepy to import CDF files - would be good to have this ability natively.

Most of them don't need to be exposed to users

Can use the example I have working locally, with the Solar Orbiter 2020 orbit (yay!).

Installation docs need to reflect:

• Installation of spacepy
• Installation of the CDF library

Try to get readthedocs working again. Probably depends on #5

Add common coordinate systems used in space physics. This should probably be based on the SunPy coordinate module (http://docs.sunpy.org/en/stable/code_ref/coordinates.html), which is based on the astropy coordinate system framework (http://docs.astropy.org/en/stable/coordinates/index.html#astropy-coordinates)

Ideas for coordinate systems to implement:

• HEE
• GSE
• RTN
• GSM

Currently https://github.com/AndrewAnnex/SpiceyPy is a really useful python package for loading spice kernels, but it would be useful to have a higher level interface in HelioPy for retrieving orbital information from spice kernels.

Currently the Python docs only have 3.5 and 3.6, so presumably just need to add a 3.5 build.

Currently all the data returned by HelioPy just returns numbers, and (hopefully!) the documentation says what units the numbers are in. AstroPy has a mature units system however, that allows data that has units attached: http://docs.astropy.org/en/stable/units/index.html

One potential problem is that pandas DataFrames do not currently work with units: pandas-dev/pandas#10349 and all of HelioPy revolves around DataFrames, so this might involve finding an alternative for DataFrames. A possible alternative is AstroPy tables: http://docs.astropy.org/en/stable/table/index.html

This would save checking for the file online each time a local file doesn't exist and remote data doesn't exist.

The main data type that is returned by heliopy, there should probably be a paragraph or so explaining how they work and linking to the pandas documentation.

http://www.sidc.be/silso/datafiles

On windows this puts in "" instead of "/", which breaks the url.

When passing probe to a data import method, it would be good if either an integer or string could be accepted where the probe can be [1,2,3,4] for example.

See http://doc.pytest.org/en/latest/example/simple.html for how to do this

Currently they take a host of integers, would be much simpler if they took a datetime.datetime, and extracted the relevant information from that.