You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is your feature request related to a problem? Please describe.
Newer versions of sgp4 allow for the direct input of Keplerian elements to describe an orbit rather than TLEs (eccentricity, inclination, argument of periapsis, right ascension of the ascending node, mean anomaly, mean motion).
Describe the solution you'd like
It would be easier to allow these as direct inputs to the instrument rather than generate a fake TLE (#8). The new syntax also allows for creating an array of timestamps rather than looping through the code.
Describe alternatives you've considered
From a design perspective, it may be easier for some to envision periapsis and apoapsis altitudes rather than eccentricity and mean motion (inverse of orbital period). Conversion routines should be added, which are either run by the user or directly run by the pysat instrument.
Additional context
preliminary tests of this workflow in #60. May replace #8 and #13
The text was updated successfully, but these errors were encountered:
Is your feature request related to a problem? Please describe.
Newer versions of sgp4 allow for the direct input of Keplerian elements to describe an orbit rather than TLEs (eccentricity, inclination, argument of periapsis, right ascension of the ascending node, mean anomaly, mean motion).
Describe the solution you'd like
It would be easier to allow these as direct inputs to the instrument rather than generate a fake TLE (#8). The new syntax also allows for creating an array of timestamps rather than looping through the code.
Describe alternatives you've considered
From a design perspective, it may be easier for some to envision periapsis and apoapsis altitudes rather than eccentricity and mean motion (inverse of orbital period). Conversion routines should be added, which are either run by the user or directly run by the pysat instrument.
Additional context
preliminary tests of this workflow in #60. May replace #8 and #13
The text was updated successfully, but these errors were encountered: