home.md

@@ -20,16 +20,20 @@ python3 run.py path/to/case.py
 ## Use
 PyPK is initialized through its API `init_pypk`. Examples can be found under the tests directory. A case file typically contains the following lines
 ```python
+# Initialize
 from pypk import init_pypk
 cfg = {...}
 solver = init_pypk(cfg)
 solver.set_matrices(m, k, q)
+# Compute flutter solution
 solver.compute()
 solver.find_flutter()
 solver.save(case_name)
 solver.plot(case_name, show=True, format=fmt)
+# Compute gradients (only available for NIPK)
+solver.compute_gradients()
 ```
-where `cfg` is a `dict` defining the configuration of the solution method, `m`, `k` and `q` are the mass, stiffness and aerodynamic force matrices provided as `numpy array`, and `case_name` and `fmt` are `str`.
+where `cfg` is a `dict` defining the configuration of the solution method, `m`, `k` and `q` are the mass, stiffness and aerodynamic force matrices provided as `numpy array`, and `case_name` and `fmt` are `str`. The variables are stored as public attributes of the `solver` and are listed in [`Solution.__init__()`](https://gitlab.uliege.be/am-dept/pypk/-/blob/master/pypk/solution.py).
 PyPK has also been interfaced with OpenMDAO for performing flutter-constrained optimization, see [OMFlut](https://gitlab.uliege.be/am-dept/omflut). Within that context, its builder must be imported using `from pypk.api_om import PypkBuilder`, then initialized with a configuration dictionary and provided to an OMFLut FlutterGroup as `FlutterGroup(..., flutter=PypkBuilder(cfg))`.
 
 The configuration dictionary contains the following entries