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
I just really quickly want to say that this package is amazing, and has been a lot of fun to mess around with!
Moving on the issue, I noticed that while trying to plot SUVI data I was getting errors on the format of the date string. I modified the date on the sample code in the solar data sample notebook so my code is as follows: g = goes_nearesttime("2024-05-14 17:00", product="SUVI-L1b-Fe131") which gave me the ValueError below:
ValueError: time data "c20241351601329.fits" doesn't match format "c%Y%j%H%M%S%f.nc", at position 0. You might want to try:
- passing format if your strings have a consistent format;
- passing format='ISO8601' if your strings are all ISO8601 but not necessarily in exactly the same format;
- passing format='mixed', and the format will be inferred for each element individually. You might want to use dayfirst alongside this.
As a test I used the sample code in the SUVI notebook: g = goes_nearesttime("2020-01-04 00:00", product="SUVI-L1b-Fe131") and had success. After some testing I was able to find that any date before August 24th 2022 works, but any date afterwards doesn't.
I'm running a jupyter notebook in VSCode using Python 3.11.9.
The text was updated successfully, but these errors were encountered:
I just really quickly want to say that this package is amazing, and has been a lot of fun to mess around with!
Moving on the issue, I noticed that while trying to plot SUVI data I was getting errors on the format of the date string. I modified the date on the sample code in the solar data sample notebook so my code is as follows:
g = goes_nearesttime("2024-05-14 17:00", product="SUVI-L1b-Fe131")
which gave me the ValueError below:ValueError: time data "c20241351601329.fits" doesn't match format "c%Y%j%H%M%S%f.nc", at position 0. You might want to try:
- passing
format
if your strings have a consistent format;- passing
format='ISO8601'
if your strings are all ISO8601 but not necessarily in exactly the same format;- passing
format='mixed'
, and the format will be inferred for each element individually. You might want to usedayfirst
alongside this.As a test I used the sample code in the SUVI notebook:
g = goes_nearesttime("2020-01-04 00:00", product="SUVI-L1b-Fe131")
and had success. After some testing I was able to find that any date before August 24th 2022 works, but any date afterwards doesn't.I'm running a jupyter notebook in VSCode using Python 3.11.9.
The text was updated successfully, but these errors were encountered: