When providing a table with time and positions to MovintTPF(), cadences can be removed during position interpolation to match the FFI cube times.
Example:
sss_asteroid_4_3_3_track_1.csv
track = pd.read_csv("sss_asteroid_4_3_3_track_1.csv")
# 19 frames in track
target = MovingTPF("Asteroid_3", track)
target.get_data(shape=(21,21))
# gets 18 frames, drops the first frame
The interpolation in:
|
column_interp = np.interp( |
A possible solution is to round the times, from the input track and FFI cube to a millisecond precision (8th decimal point in days). TESS time is precise to 0.1 seconds. This in truth changes following definitions of positions based on trajectories and time, but it is negligible at a millisecond scale.
When providing a table with time and positions to
MovintTPF(),cadences can be removed during position interpolation to match the FFI cube times.Example:
sss_asteroid_4_3_3_track_1.csv
The interpolation in:
tess-asteroids/src/tess_asteroids/movingtpf.py
Line 139 in 08c4123
A possible solution is to round the times, from the input track and FFI cube to a millisecond precision (8th decimal point in days). TESS time is precise to 0.1 seconds. This in truth changes following definitions of positions based on trajectories and time, but it is negligible at a millisecond scale.