PCDS conda Python environments already include archapp.
In order to access the archiver interactively, use the following:
Source the environment first:
$ source /cds/group/pcds/pyps/conda/pcds_conda
And then start IPython:
$ ipython
Import a helper class that lets you access the archiver:
from archapp.interactive import EpicsArchive
Now create a reusable object that will let you access our photon-side archiver:
arch = EpicsArchive()
To get data, use the following:
data = arch.get("pv_name", xarray=True)
It's recommended to use xarray=True
in order to get an easy-to-use time series
array back.
To get the data as a pandas data frame, you can use:
df = data.to_dataframe()
To get just values (or alarm status) from the data frame, use the following:
df = data.to_dataframe()
values = df["pv_name"].vals
alarm_stat = df["pv_name"].stat
To resample the values to just once per day, interpolating linearly (common alternative options may include "quadratic" or "cubic" or "nearest"):
data.resample(time="1D").interpolate("linear")
To merge data from different PVs - using a real pair for example - one option may be the following:
In [1]: m1 = arch.get("AT2K2:L2SI:MMS:01.RBV", xarray=True).resample(time="1D").interpolate("linear").to_dataframe()
In [2]: m2 = arch.get("AT2K2:L2SI:MMS:02.RBV", xarray=True).resample(time="1D").interpolate("linear").to_dataframe()
In [3]: m1.merge(m2, left_index=True, right_index=True)[:10]
Out[3]:
AT2K2:L2SI:MMS:01.RBV AT2K2:L2SI:MMS:02.RBV
field time
vals 2021-08-09 26.994 27.71345
2021-08-10 26.994 27.71345
2021-08-11 26.9939 27.71345
2021-08-12 26.9939 27.71345
2021-08-13 26.9939 27.7134
2021-08-14 26.9939 27.71325
2021-08-15 26.9939 27.71335
2021-08-16 26.99395 27.71335
2021-08-17 26.994 27.7135
2021-08-18 26.994 27.71345
Create a conda environment that uses python 3. Make sure xarray is installed in the conda environment.
Go to archapp/lib and type ipython
in the command line.