Currently obtaining data for an Obspack time series will return a pandas dataframe with a time column.
Example code:
from icoscp.cpb.dobj import Dobj
dobj = Dobj('https://meta.icos-cp.eu/objects/VvC1QkMAWW5Z9iUSFaPijY5W')
dobj.data
There are other time-related columns that should be in the dataframe but are not.
Here's the code that does the conversions:
|
""" |
|
The ICOS Carbon Portal provides a TIMESTAMP which is |
|
a unix timestamp in milliseconds [UTC] |
|
Convert the "number" to a pandas date/time object |
|
""" |
|
|
|
if 'TIMESTAMP' in df.columns and self._dtconvert: |
|
df['TIMESTAMP'] = pd.to_datetime(df.loc[:,'TIMESTAMP'],unit='ms') |
|
|
|
if 'TIMESTAMP_END' in df.columns and self._dtconvert: |
|
df['TIMESTAMP_END'] = pd.to_datetime(df.loc[:,'TIMESTAMP_END'],unit='ms') |
|
|
|
# there are files with date and time where |
|
# date is a unixtimestamp |
|
# time seconds per day |
|
if 'date' in df.columns and self._dtconvert: |
|
df['date'] = pd.to_datetime(df.loc[:,'date'],unit='D') |
|
|
|
if 'time' in df.columns and self._dtconvert: |
|
# convert unix timestamp |
|
df['time'] = pd.to_datetime(df.loc[:,'time'],unit='s') |
|
# remove the data, so that only time remains |
|
df['time'] = pd.to_datetime(df.loc[:,'time'],format='%H:%M').dt.time |
