Using the duqmap function¶
The api provides a map function duqmap which can be used to map functions over data generated by duqtools, or just data that you specify yourself (as in this example)
In [1]:
Copied!
import xarray as xr
from duqtools.api import ImasHandle, Run, Runs, duqmap, rebase_all_coords
import xarray as xr
from duqtools.api import ImasHandle, Run, Runs, duqmap, rebase_all_coords
Read data from runs.yaml¶
This data is later used to instrument duqtools over which runs to apply the function
In [2]:
Copied!
runs = Runs.parse_file("example/runs.yaml")
for run in runs:
run.data_out.relative_location = None
runs = Runs.parse_file("example/runs.yaml")
for run in runs:
run.data_out.relative_location = None
Define a function to be used with duqmap¶
Functions should have either Run or ImasHandle as the type of the first argument, duqmap will then take care of converting the input to the correct type
In [3]:
Copied!
def print_and_return_handle(run: Run):
print(run.dirname)
return str(run.dirname)
def print_and_return_handle(run: Run):
print(run.dirname)
return str(run.dirname)
this function extracts the dirnames from any runs duqmap operates over:
In [4]:
Copied!
names = duqmap(print_and_return_handle, runs=runs)
names = duqmap(print_and_return_handle, runs=runs)
run_0000 run_0001 run_0002
A function to extract the electron temperature of each ImasHandle for each output mapped over by duqmap:
In [5]:
Copied!
def extract_te(handle: ImasHandle):
return handle.get_variables(['t_e', 'rho_tor_norm', 'time'])
def extract_te(handle: ImasHandle):
return handle.get_variables(['t_e', 'rho_tor_norm', 'time'])
Retrieve a dataset for each of the runs specified:
In [6]:
Copied!
datasets = duqmap(extract_te, runs=runs)
datasets
datasets = duqmap(extract_te, runs=runs)
datasets
Out[6]:
[<xarray.Dataset>
Dimensions: (time: 1, rho_tor_norm: 101)
Coordinates:
* time (time) float64 45.75
* rho_tor_norm (rho_tor_norm) float64 0.0 0.01 0.02 0.03 ... 0.98 0.99 1.0
Data variables:
t_e (time, rho_tor_norm) float64 4.883e+03 4.881e+03 ... 218.1,
<xarray.Dataset>
Dimensions: (time: 1, rho_tor_norm: 101)
Coordinates:
* time (time) float64 45.75
* rho_tor_norm (rho_tor_norm) float64 0.0 0.01 0.02 0.03 ... 0.98 0.99 1.0
Data variables:
t_e (time, rho_tor_norm) float64 4.131e+03 4.13e+03 ... 184.5,
<xarray.Dataset>
Dimensions: (time: 1, rho_tor_norm: 101)
Coordinates:
* time (time) float64 45.75
* rho_tor_norm (rho_tor_norm) float64 0.0 0.01 0.02 0.03 ... 0.98 0.99 1.0
Data variables:
t_e (time, rho_tor_norm) float64 4.507e+03 4.505e+03 ... 201.3]
Rebase and convert all the data into a single dataset with the help of xarray
In [7]:
Copied!
datasets = rebase_all_coords(datasets, datasets[0])
dataset = xr.concat(datasets, dim='run')
dataset['run'] = names
dataset
datasets = rebase_all_coords(datasets, datasets[0])
dataset = xr.concat(datasets, dim='run')
dataset['run'] = names
dataset
Out[7]:
<xarray.Dataset>
Dimensions: (run: 3, time: 1, rho_tor_norm: 101)
Coordinates:
* time (time) float64 45.75
* rho_tor_norm (rho_tor_norm) float64 0.0 0.01 0.02 0.03 ... 0.98 0.99 1.0
* run (run) <U8 'run_0000' 'run_0001' 'run_0002'
Data variables:
t_e (run, time, rho_tor_norm) float64 4.883e+03 ... 201.3Plot our results the way we want¶
In [8]:
Copied!
dataset['t_e'].isel(time=0).plot.line(x='rho_tor_norm', hue='run')
dataset['t_e'].isel(time=0).plot.line(x='rho_tor_norm', hue='run')
Out[8]:
[<matplotlib.lines.Line2D at 0x7fdae8cb6f70>, <matplotlib.lines.Line2D at 0x7fdae8cb6fa0>, <matplotlib.lines.Line2D at 0x7fdae8ccd040>]
In [ ]:
Copied!