Working with IMAS files
This example shows how duqtools can be used to manipulate IMAS data using ImasHandle and IDSMapping.
Copying data
This example shows how duqtools can be used to copy data.
from duqtools.api import ImasHandle
First, we construct and Imas handle, consisting of the user name, database (sometimes referred to as machine name), shot and run number. Together these form an Imas path that point to some data on the system.
We can print the path, and check that the data exists.
Note that the user is optional. Duqtools will fill in the current user by default.
handle = ImasHandle(user='g2ssmee', db='jet', shot=94875, run=8000)
print(handle.path())
print(handle.exists())
Let's say we want to make a copy of this data set to modify. We should construct a new Imas handle to specify the target location. In the next cell, we create the Imas handle from a string instead. Again, the user is optional, so the string 'jet/94875/8888' would give the same result.
target = ImasHandle.from_string('g2ssmee/jet/94785/8888')
print(target.path())
print(target.exists())
The template data can be copied to the new location.
handle.copy_data_to(target)
print(target.exists())
If you are unhappy with the outcome, wecan delete the data using the ImasHandle.delete() method. Let's not do that for now.
# target.delete()
print(target.exists())
Local IMAS databases
Duqtools is compatible with local imas databases, so you can specify the path to your imasdb directly:
handle = ImasHandle(user='/afs/eufus.eu/user/g/g2ssmee/public/imasdb',
db='jet',
shot=94875,
run=8000)
print(handle.path())
print(handle.exists())
Exploring, modifying, and saving data
We developed duqtools to make it as straightforward as possible to manipulate IDSs. In this section, we show an example of how one may go about doing so.
Let's say we want to look at 'core_profiles' > 'profiles_1d' > 't_i_average'.
First, we retrieve the 'core_profiles' IDS.
cp = target.get('core_profiles')
print(cp)
Duqtools wraps the IMAS database in a dict-like object, so that the different datasets can be accessed like a dict.
cp['profiles_1d/0/t_i_average']
cp['profiles_1d/0/t_i_average'] *= 1.1
Once we changed the data, we can sync the changes back to the data file.
target.update_from(cp)
Plotting data
The example below shows how to plot some data using matplotlib.
x = cp['profiles_1d/0/grid/rho_tor_norm']
y = cp['profiles_1d/0/t_i_average']
import matplotlib.pyplot as plt
fig, ax = plt.subplots()
ax.plot(x, y)
ax.set_xlabel('rho_tor_norm')
ax.set_ylabel('t_i_average');
