Solar wind in situ data suitable for machine learning (python numpy structured arrays): STEREO-A/B, Wind, Parker Solar Probe, Ulysses, Venus Express, MESSENGER

These are solar wind in situ data arrays in python pickle format suitable for machine learning, i.e. the arrays consist only of numbers, no strings and no datetime objects.

See AAREADME_insitu_ML.txt for more explanation.

If you use these data for peer reviewed scientific publications, please get in touch concerning usage and possible co-authorship by the authors (C. Möstl, A. J. Weiss, R. L. Bailey, A. Isavnin): christian.moestl@oeaw.ac.at or twitter @chrisoutofspace

Made with https://github.com/cmoestl/heliocats

Load in python with e.g. for Parker Solar Probe data:

> import pickle

> filepsp='psp_2018_2019_sceq_ndarray.p'

> [psp,hpsp]=pickle.load(open(filepsp, "rb" ) )

plot time vs total field

> import matplotlib.pyplot as plt

> plt.plot(psp['time'],psp['bt'])

Times psp[:,0 ] or psp['time'] are in matplotlib format.

Variable 'hpsp' contains a header with the variable names and units for each column. Coordinate systems for magnetic field components are RTN (Ulysses), SCEQ (Parker Solar Probe, STEREO-A/B, VEX, MESSENGER), HEEQ (Wind)

available parameters:

bt = total magnetic field

bxyz = magnetic field components

vt = total proton speed

vxyz = velocity components (only for PSP)

np = proton density

tp = proton temperature

xyz = spacecraft position in HEEQ

r, lat, lon = spherical coordinates of position in HEEQ