Typo in Antares astro Benchmark moon and sun sample
Looking at the Antares astro benchmark for the moon and sun position, I noticed something potentially off.
It can be reproduced using this small code snippet:
import numpy as np
import pandas as pd
from km3net_testdata import data_path
csv_file = data_path("astro/antares_moon_sun_position_benchmark.csv")
df = pd.read_csv(csv_file,header=8)
df = df.set_index('date')
print (df)
print ('Difference between RA and RA-J2000')
print (df['RA'] - df['RA-J2000'])
The output is the following:
date time object zenith azimuth DEC RA DEC-J2000 RA-J2000
0 2008-11-08 23:34:54.00 Moon 69.498 201.331 -1.914 348.977 -1.963 348.860
1 2008-10-22 10:30:45.40 Moon 57.834 184.180 17.197 133.306 17.230 133.179
2 2002-03-02 18:20:35.30 Moon 122.894 19.697 -8.162 209.628 -8.153 209.604
3 2000-11-15 07:03:22.50 Moon 59.308 175.671 21.917 98.854 21.918 98.846
4 2008-11-08 23:34:54.00 Sun 153.882 79.495 -16.897 224.474 -16.860 224.348
5 2008-10-22 10:30:45.40 Sun 55.210 282.752 -11.260 201.331 -11.215 207.211
6 2002-03-02 18:20:35.30 Sun 100.873 177.667 -7.067 343.387 -7.077 343.363
7 2000-11-15 07:03:22.50 Sun 85.016 326.969 -18.579 230.829 -18.578 230.822
Difference between RA and RA-J2000
2008-11-08 0.117
2008-10-22 0.127
2002-03-02 0.024
2000-11-15 0.008
2008-11-08 0.126
2008-10-22 -5.880
2002-03-02 0.024
2000-11-15 0.007
dtype: float64
In the first table, we can see the coordinates in the RA/DEC system, with or without the J2000 epoch.
The second table show the difference between RA and RA-J2000. As expected, the difference between both is quite small, below a degree. We can also observe that the shift depends on the date. However, one row is off: 2008-10-22 -5.880
, by almost 6 degree.
This correspond to this row in the first table:
date time object zenith azimuth DEC RA DEC-J2000 RA-J2000
5 2008-10-22 10:30:45.40 Sun 55.210 282.752 -11.260 201.331 -11.215 207.211
Looking at this line, I suspect that 201.331
should be 207.331
and that somehow a typo was done, swapping the 7 by a 1.
This potential typo is also present in http://antares-wiki.in2p3.fr/doku.php?id=benchmarks_astro
I was using this sample to benchmark a custom code I was writing, which seems to indicate that the RA-J2000 coordinate is the proper one.
Maybe @dsamtleben or @thomas.eberl can confirm this ?