dynamicslab / pykoopman Goto Github PK

View Code? Open in Web Editor NEW

252.0 252.0 50.0 41.33 MB

A package for computing data-driven approximations to the Koopman operator.

Home Page: https://pykoopman.readthedocs.io

License: MIT License

Python 100.00%

pykoopman's People

Contributors

Stargazers

Watchers

pykoopman's Issues

Example not working after installation

I installed the package using pip, and am trying to run the tutorial_koopman_edmdc example. However, I get error when trying to import as shown below:

ImportError Traceback (most recent call last)
in
1 get_ipython().run_line_magic('matplotlib', 'inline')
2 import pykoopman as pk
----> 3 from pykoopman.common.examples import vdp_osc, rk4, square_wave # required for example system
4 import matplotlib.pyplot as plt
5 import numpy as np

ImportError: cannot import name 'vdp_osc' from 'pykoopman.common.examples' (/Users/jehanzeb/anaconda3/lib/python3.8/site-packages/pykoopman/common/examples.py)

How to save and load the regressor or model?

Hi,

I am trying to implement the NNDMD and want to save the learned dlk_regressor for future use. How could I save and load the learned regressors and models (e.g., run with pk.Koopman.fit())?

Thanks in advance for any assistance.

Bug in model.fit with timedelays

from pykoopman.regression import EDMD
from sklearn.metrics import mean_absolute_percentage_error
from sklearn.metrics import mean_absolute_error
pred = 4
S = np.array(data)
n_output_vars = S.shape[1]  # Get the number of output variables
n_delays = 100
n = 310
regressor = EDMD()

# Loop through each timestep
for timestep in range(n, len(S) - pred):
    try:
        obs = TimeDelay(n_delays=n_delays)

        # Prepare input data for the current timestep
        X1 = S[:timestep-1].T
        X2 = S[1:timestep].T

        model = pk.Koopman(observables=obs, regressor=regressor)
        model.fit(X1.T, y=X2.T) 

        n_steps = timestep - n_delays + pred
        x0_td = X1[:,:n_delays+1].T

        Xkoop = model.simulate(x0_td, n_steps=n_steps)
        Xkoop2 = np.vstack([x0_td, Xkoop]) # add initial condition to simulated data for comparison below

    except ValueError as ve:
        print("Timestep:", timestep, "X1 shape:", X1.shape)
        print(ve)
        continue
        ```
        
```Timestep: 401 X1 shape: (488, 400)
x has too few rows. To compute time-delay features with delay = 1 and n_delays = 100 x must have at least 101 rows.```

It happens at timestep 301 too.
All other timesteps complete fine. 

99 time delays seems fine. Delays >100 error.

`python setup.py install` doesn't work

pykoopman/README.rst

Line 173 in a3c6f98

python setup.py install

the file setup.py doesn't exist

changing the activation function

When using NNDMD, it seems that changing the activation function results in an error. Could this be because the library does not define activation functions other than linear?

NAN issues

I encounter NaN (Not a Number) issues when training certain specific functions through the network. There are no NaN values in my training set, so I suspect it's due to gradient explosion. I'm unsure how to resolve this problem

Documentation typo

I believe that in the HAVOK/Lorenz example, in cell 9 under "Prediction using the HAVOK model", we should be plotting xpred[:,j], not xpred[:,i]. This (very minor) bug doesn't actually show up because it so happens that since INDEX is [0], [1], [2], or [1,2], we always have i == j. However, if we were to choose INDEX = [0, 2], then the bug would appear.

Changing time step in `simulate`

Hi,
first thank you for the very nice package!
If I understand it correctly, Koopman.simulate simulates the learned model forward in time with the same time interval dt as the one which was used to fit the model, right ? If so, would it be possible to include an option to change the dt for this prediction. For instance, I would like to test the quality of the learning with a small number of snapshots over a given time interval and check against the solution on the intermediate time steps between 2 snapshots used for the training.
Thanks!

Question about the difference between the function named '_compute_phi' and '_compute_psi'

Take the _edmd.py in pykoopman.regression as an example. In this file, phi is computed as self._ur.conj().T @ x_col, while psi is computed as self._tmp_compute_psi @ x_col. I can't figure out the true meaning behind the phi and psi, not to mention their difference. Could you please explain it in more detail by Mathematical formula? It would be more appreciated if there are corresponding articles to provide.

koopman.py erros

Hello,

While running example: tutorial_eigenfunctions_with_EDMD.ipynb, I ran into error in cell 5.

Apparently, in koopman.py, line 562: return self.model.steps[-1][1].kef_

"kef_" was not defined. Also see below. Please help! Thanks!

AttributeError Traceback (most recent call last)
Cell In[5], line 7
4 model.fit(X.T, y=Y.T)
6 print(model.eigenvalues)
----> 7 sys.visualize_modes(X, model.eigenfunctions)

File /pykoopman/koopman.py:562, in Koopman.eigenfunctions(self)
560 check_is_fitted(self, "model")
561 #return self.model.steps[-1][1].eigenfunctions_
--> 562 return self.model.steps[-1][1].kef_

AttributeError: 'EDMD' object has no attribute 'kef_'

The predicted derivative.

Is there a function that can predict the derivative with respect to time of a certain state variable?

Running "Neural Network DMD on Slow Manifold" error

I have lightning version 2.0.5 installed. The pykoopman version is 1.0.4.
When I run "dlk_regressor.fit(traj_list)" in the example "tutorial_koopman_nndmd_examples.ipynb".

It shows errors

INFO: GPU available: True (cuda), used: True
[rank_zero.py:48 -                _info() ] GPU available: True (cuda), used: True
INFO: TPU available: False, using: 0 TPU cores
[rank_zero.py:48 -                _info() ] TPU available: False, using: 0 TPU cores
INFO: IPU available: False, using: 0 IPUs
[rank_zero.py:48 -                _info() ] IPU available: False, using: 0 IPUs
INFO: HPU available: False, using: 0 HPUs
[rank_zero.py:48 -                _info() ] HPU available: False, using: 0 HPUs
INFO: Initializing distributed: GLOBAL_RANK: 1, MEMBER: 2[/2](https://file+.vscode-resource.vscode-cdn.net/2)
[distributed.py:245 - _init_dist_connection() ] Initializing distributed: GLOBAL_RANK: 1, MEMBER: 2[/2](https://file+.vscode-resource.vscode-cdn.net/2)
INFO: Initializing distributed: GLOBAL_RANK: 0, MEMBER: 1[/2](https://file+.vscode-resource.vscode-cdn.net/2)
[distributed.py:245 - _init_dist_connection() ] Initializing distributed: GLOBAL_RANK: 0, MEMBER: 1[/2](https://file+.vscode-resource.vscode-cdn.net/2)
INFO: ----------------------------------------------------------------------------------------------------
distributed_backend=nccl
All distributed processes registered. Starting with 2 processes
----------------------------------------------------------------------------------------------------

[rank_zero.py:48 -                _info() ] ----------------------------------------------------------------------------------------------------
distributed_backend=nccl
All distributed processes registered. Starting with 2 processes
----------------------------------------------------------------------------------------------------

2023-09-14 16:20:30.863437: I tensorflow[/core/platform/cpu_feature_guard.cc:193](https://file+.vscode-resource.vscode-cdn.net/core/platform/cpu_feature_guard.cc:193)] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 AVX512F AVX512_VNNI FMA
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2023-09-14 16:20:31.016351: I tensorflow[/core/util/port.cc:104](https://file+.vscode-resource.vscode-cdn.net/core/util/port.cc:104)] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
2023-09-14 16:20:32.029631: W tensorflow[/compiler/xla/stream_executor/platform/default/dso_loader.cc:64](https://file+.vscode-resource.vscode-cdn.net/compiler/xla/stream_executor/platform/default/dso_loader.cc:64)] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: [/usr/local/cuda-11.7/lib64](https://file+.vscode-resource.vscode-cdn.net/usr/local/cuda-11.7/lib64)
2023-09-14 16:20:32.029742: W tensorflow[/compiler/xla/stream_executor/platform/default/dso_loader.cc:64](https://file+.vscode-resource.vscode-cdn.net/compiler/xla/stream_executor/platform/default/dso_loader.cc:64)] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: [/usr/local/cuda-11.7/lib64](https://file+.vscode-resource.vscode-cdn.net/usr/local/cuda-11.7/lib64)
2023-09-14 16:20:32.029751: W tensorflow[/compiler/tf2tensorrt/utils/py_utils.cc:38](https://file+.vscode-resource.vscode-cdn.net/compiler/tf2tensorrt/utils/py_utils.cc:38)] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
---------------------------------------------------------------------------
ProcessRaisedException                    Traceback (most recent call last)
Cell In[6], line 1
----> 1 dlk_regressor.fit(traj_list)

File [~/anaconda3/envs/dpc/lib/python3.10/site-packages/pykoopman/regression/_nndmd.py:1187](https://file+.vscode-resource.vscode-cdn.net/media/lk/lksgcc/lk_git/3_Reinforcement_Learning/3_4_MPC/pykoopman/docs/~/anaconda3/envs/dpc/lib/python3.10/site-packages/pykoopman/regression/_nndmd.py:1187), in NNDMD.fit(self, x, y, dt)
   1184     raise ValueError("check `x` and `y` for `self.fit`")
   1186 # trainer starts to train
-> 1187 self.trainer.fit(self._regressor, self.dm)
   1189 # compute Koopman operator information
   1190 self._state_matrix_ = (
   1191     self._regressor._koopman_propagator.get_discrete_time_Koopman_Operator()
   1192     .detach()
   1193     .numpy()
   1194 )

File [~/anaconda3/envs/dpc/lib/python3.10/site-packages/lightning/pytorch/trainer/trainer.py:529](https://file+.vscode-resource.vscode-cdn.net/media/lk/lksgcc/lk_git/3_Reinforcement_Learning/3_4_MPC/pykoopman/docs/~/anaconda3/envs/dpc/lib/python3.10/site-packages/lightning/pytorch/trainer/trainer.py:529), in Trainer.fit(self, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path)
    527 model = _maybe_unwrap_optimized(model)
    528 self.strategy._lightning_module = model
--> 529 call._call_and_handle_interrupt(
    530     self, self._fit_impl, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path
    531 )

File [~/anaconda3/envs/dpc/lib/python3.10/site-packages/lightning/pytorch/trainer/call.py:41](https://file+.vscode-resource.vscode-cdn.net/media/lk/lksgcc/lk_git/3_Reinforcement_Learning/3_4_MPC/pykoopman/docs/~/anaconda3/envs/dpc/lib/python3.10/site-packages/lightning/pytorch/trainer/call.py:41), in _call_and_handle_interrupt(trainer, trainer_fn, *args, **kwargs)
     39 try:
     40     if trainer.strategy.launcher is not None:
---> 41         return trainer.strategy.launcher.launch(trainer_fn, *args, trainer=trainer, **kwargs)
     42     return trainer_fn(*args, **kwargs)
     44 except _TunerExitException:

File [~/anaconda3/envs/dpc/lib/python3.10/site-packages/lightning/pytorch/strategies/launchers/multiprocessing.py:124](https://file+.vscode-resource.vscode-cdn.net/media/lk/lksgcc/lk_git/3_Reinforcement_Learning/3_4_MPC/pykoopman/docs/~/anaconda3/envs/dpc/lib/python3.10/site-packages/lightning/pytorch/strategies/launchers/multiprocessing.py:124), in _MultiProcessingLauncher.launch(self, function, trainer, *args, **kwargs)
    116 process_context = mp.start_processes(
    117     self._wrapping_function,
    118     args=process_args,
   (...)
    121     join=False,  # we will join ourselves to get the process references
    122 )
    123 self.procs = process_context.processes
--> 124 while not process_context.join():
    125     pass
    127 worker_output = return_queue.get()

File [~/anaconda3/envs/dpc/lib/python3.10/site-packages/torch/multiprocessing/spawn.py:160](https://file+.vscode-resource.vscode-cdn.net/media/lk/lksgcc/lk_git/3_Reinforcement_Learning/3_4_MPC/pykoopman/docs/~/anaconda3/envs/dpc/lib/python3.10/site-packages/torch/multiprocessing/spawn.py:160), in ProcessContext.join(self, timeout)
    158 msg = "\n\n-- Process %d terminated with the following error:\n" % error_index
    159 msg += original_trace
--> 160 raise ProcessRaisedException(msg, error_index, failed_process.pid)

ProcessRaisedException: 

-- Process 1 terminated with the following error:
Traceback (most recent call last):
  File "[/home/lk/anaconda3/envs/dpc/lib/python3.10/site-packages/torch/multiprocessing/spawn.py](https://file+.vscode-resource.vscode-cdn.net/home/lk/anaconda3/envs/dpc/lib/python3.10/site-packages/torch/multiprocessing/spawn.py)", line 69, in _wrap
    fn(i, *args)
  File "[/home/lk/anaconda3/envs/dpc/lib/python3.10/site-packages/lightning/pytorch/strategies/launchers/multiprocessing.py](https://file+.vscode-resource.vscode-cdn.net/home/lk/anaconda3/envs/dpc/lib/python3.10/site-packages/lightning/pytorch/strategies/launchers/multiprocessing.py)", line 147, in _wrapping_function
    results = function(*args, **kwargs)
  File "[/home/lk/anaconda3/envs/dpc/lib/python3.10/site-packages/lightning/pytorch/trainer/trainer.py](https://file+.vscode-resource.vscode-cdn.net/home/lk/anaconda3/envs/dpc/lib/python3.10/site-packages/lightning/pytorch/trainer/trainer.py)", line 568, in _fit_impl
    self._run(model, ckpt_path=ckpt_path)
  File "[/home/lk/anaconda3/envs/dpc/lib/python3.10/site-packages/lightning/pytorch/trainer/trainer.py](https://file+.vscode-resource.vscode-cdn.net/home/lk/anaconda3/envs/dpc/lib/python3.10/site-packages/lightning/pytorch/trainer/trainer.py)", line 934, in _run
    call._call_setup_hook(self)  # allow user to setup lightning_module in accelerator environment
  File "[/home/lk/anaconda3/envs/dpc/lib/python3.10/site-packages/lightning/pytorch/trainer/call.py](https://file+.vscode-resource.vscode-cdn.net/home/lk/anaconda3/envs/dpc/lib/python3.10/site-packages/lightning/pytorch/trainer/call.py)", line 83, in _call_setup_hook
    _call_lightning_datamodule_hook(trainer, "setup", stage=fn)
  File "[/home/lk/anaconda3/envs/dpc/lib/python3.10/site-packages/lightning/pytorch/trainer/call.py](https://file+.vscode-resource.vscode-cdn.net/home/lk/anaconda3/envs/dpc/lib/python3.10/site-packages/lightning/pytorch/trainer/call.py)", line 164, in _call_lightning_datamodule_hook
    return fn(*args, **kwargs)
  File "[/home/lk/anaconda3/envs/dpc/lib/python3.10/site-packages/pykoopman/regression/_nndmd.py](https://file+.vscode-resource.vscode-cdn.net/home/lk/anaconda3/envs/dpc/lib/python3.10/site-packages/pykoopman/regression/_nndmd.py)", line 902, in setup
    self._tr_x, self._tr_yseq, self._tr_ys, self.normalization
AttributeError: 'SeqDataModule' object has no attribute '_tr_x'

Anyone can help ?

no version ranges for dependencies

pykoopman/requirements.txt

Line 2 in a3c6f98

derivative

time delays not working correctly

tutorial_koopman_hankel_dmdc_for_vdp_system.ipynb

%matplotlib inline
import pykoopman as pk
from pykoopman.common.examples import vdp_osc, rk4, square_wave  # required for example system
import matplotlib.pyplot as plt
import numpy as np
import numpy.random as rnd
np.random.seed(42)  # for reproducibility

import warnings
warnings.filterwarnings('ignore')

n_states = 2 # Number of states
n_inputs = 1 # Number of control inputs
dT = 0.01    # Timestep
n_traj = 200  # Number of trajectories
n_int = 1000  # Integration length
# Time vector
t = np.arange(0, n_int*dT, dT)

# Uniform random distributed forcing in [-1, 1]
u = 2*rnd.random([n_int, n_traj])-1

# Uniform distribution of initial conditions
x = 2*rnd.random([n_states, n_traj])-1

# Init
X = np.zeros((n_states, n_int*n_traj))
Y = np.zeros((n_states, n_int*n_traj))
U = np.zeros((n_inputs, n_int*n_traj))

# Integrate
for step in range(n_int):
    y = rk4(0, x, u[step, :], dT, vdp_osc)
    X[:, (step)*n_traj:(step+1)*n_traj] = x
    Y[:, (step)*n_traj:(step+1)*n_traj] = y
    U[:, (step)*n_traj:(step+1)*n_traj] = u[step, :]
    x = y
# Visualize first 100 steps of the training data
fig, axs = plt.subplots(1, 1, tight_layout=True, figsize=(12, 4))
for traj_idx in range(n_traj):
    x = X[:, traj_idx::n_traj]
    axs.plot(t[0:100], x[1, 0:100], 'k')
axs.set(
        ylabel=r'$x_2$',
        xlabel=r'$t$')

EDMDc = pk.regression.DMDc()

n_delays = 9
obs = pk.observables.TimeDelay(n_delays=n_delays)

model = pk.Koopman(observables=obs, regressor=EDMDc)
model.fit(X.T, y=Y.T, u=U[:,n_delays:].T)
Traceback (most recent call last):

  Cell In[291], line 53
    model.fit(X.T, y=Y.T, u=U[:,n_delays:].T)

  File ~\anaconda3\envs\DMD\lib\site-packages\pykoopman\koopman.py:153 in fit
    self.model.fit(x, y, regressor__u=u, regressor__dt=dt)

  File ~\anaconda3\envs\DMD\lib\site-packages\sklearn\pipeline.py:394 in fit
    self._final_estimator.fit(Xt, y, **fit_params_last_step)

  File ~\anaconda3\envs\DMD\lib\site-packages\sklearn\compose\_target.py:229 in fit
    self._fit_transformer(y_2d)

  File ~\anaconda3\envs\DMD\lib\site-packages\sklearn\compose\_target.py:175 in _fit_transformer
    self.transformer_.fit(y)

  File ~\anaconda3\envs\DMD\lib\site-packages\sklearn\preprocessing\_function_transformer.py:165 in fit
    self._check_inverse_transform(X)

  File ~\anaconda3\envs\DMD\lib\site-packages\sklearn\preprocessing\_function_transformer.py:136 in _check_inverse_transform
    if not _allclose_dense_sparse(X[idx_selected], X_round_trip):

  File ~\anaconda3\envs\DMD\lib\site-packages\sklearn\utils\validation.py:1607 in _allclose_dense_sparse
    return np.allclose(x, y, rtol=rtol, atol=atol)

  File <__array_function__ internals>:180 in allclose

  File ~\anaconda3\envs\DMD\lib\site-packages\numpy\core\numeric.py:2265 in allclose
    res = all(isclose(a, b, rtol=rtol, atol=atol, equal_nan=equal_nan))

  File <__array_function__ internals>:180 in isclose

  File ~\anaconda3\envs\DMD\lib\site-packages\numpy\core\numeric.py:2375 in isclose
    return within_tol(x, y, atol, rtol)

  File ~\anaconda3\envs\DMD\lib\site-packages\numpy\core\numeric.py:2356 in within_tol
    return less_equal(abs(x-y), atol + rtol * abs(y))

ValueError: operands could not be broadcast together with shapes (100,2) (91,2)

I tried also normal EDMD with time delay as an observable and it errored with the same message.

EDMD = pk.regression.EDMD()
model = pk.Koopman(observables=obs, regressor=EDMD)
model.fit(X.T, y=Y.T)

No module named optht

After installing pykoopman, I ran the corresponding program (import pykoopman as pk in Python 3.10.9) and found the following problem:
ModuleNotFoundError: No module named 'optht'

You must include `torch` in your setup.py (and `requirements.txt`) if you import torch in the pykoopman package

pykoopman/pykoopman/regression/_nndmd.py

Line 10 in a3c6f98

import torch

Other norms for fitting? (feature request)

Hi thank you for making this awesome package,

I understand that several works have noted that the L2 norm may not be the best choice for fitting certain functional spaces. Personally, I have an application where the maximum error (Linf) is more important than the mean error (L2). One may also want a sparse fit in some cases too or a lasso regression. Would you consider adding minimization (via linear programming) over other norms as a possibility?

I see this package currently is using pinv and avoiding optimization, but perhaps code from pysindy might make this feasible?

Array shape issues with non-identity observables

Currently non-identity observables have problems handling one-dimensional input vectors. validate_input reshapes them into (n,1) arrays, but this isn't always the desired behavior: sometimes the array elements are different features rather than different samples, in which case the array should have shape (1,n). Usually you can fix this by just reshaping your array to be (1,n) to begin with -- except when running running simulate in koopman.py. Even if the x0 you input has shape (1,n), the first call to predict will work as expected, but subsequent iterations will take one-dimensional slices of the array y, which are then reshaped to (n,1) arrays by validate_input leading to errors.

My current "fix" is to simply replace the line y[k + 1] = self.predict(y[k]) in simulate with y[k + 1] = self.predict(y[k].reshape(1, -1)). This passes the unit tests and seems to allow predict to work for non-identity observables, but I don't understand this code nearly well enough to say whether this is the right way to fix this or if this will cause other issues. (I think a similar fix works for when u is not None: y[k + 1] = self.predict(y[k].reshape(1,-1), u[k + 1].reshape(1,-1)), but I'm even less confident here.) It might also be better to do the reshaping in some other part of the code rather than here. Different parts of the code seem to depend on various things being specific shapes in ways that I don't understand, so it's entirely possible that this will cause problems elsewhere.

I hope this helps!

`pip install .` does not install dependencies

after running pip install . the dependencies mentionend in the requirements.txt are not installed

AttributeError: 'Koopman' object has no attribute 'V'

In the tutorial “Dynamic mode decomposition with control (DMDc) for a 128D system”，when i run the ：
DMDc = pk.regression.DMDc(svd_rank=11, svd_output_rank=10)
model = pk.Koopman(regressor=DMDc)
model.fit(torus.X.T, u=U[:,np.newaxis], dt=dt)
Phi_dmdc = model.V
torus.viz_all_modes(np.real(Phi_dmdc[:,::2]))
The program will report an error, displaying the error of “AttributeError: 'Koopman' object has no attribute 'V'”.

Installation issue with python 3.12

Fresh env. Tried both pypi and source installation: Got the following error:

Collecting scikit-learn<=1.1.3,>=1.1.3 (from pykoopman==1.0.8)
  Using cached scikit-learn-1.1.3.tar.gz (7.0 MB)
  Installing build dependencies ... done
  Getting requirements to build wheel ... error
  error: subprocess-exited-with-error
  
  × Getting requirements to build wheel did not run successfully.
  │ exit code: 1
  ╰─> [33 lines of output]
      Traceback (most recent call last):
        File "anaconda3/envs/koopman/lib/python3.12/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 353, in <module>
          main()
        File "anaconda3/envs/koopman/lib/python3.12/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 335, in main
          json_out['return_val'] = hook(**hook_input['kwargs'])
                                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        File "anaconda3/envs/koopman/lib/python3.12/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 112, in get_requires_for_build_wheel
          backend = _build_backend()
                    ^^^^^^^^^^^^^^^^
        File "anaconda3/envs/koopman/lib/python3.12/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 77, in _build_backend
          obj = import_module(mod_path)
                ^^^^^^^^^^^^^^^^^^^^^^^
        File "anaconda3/envs/koopman/lib/python3.12/importlib/__init__.py", line 90, in import_module
          return _bootstrap._gcd_import(name[level:], package, level)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        File "<frozen importlib._bootstrap>", line 1381, in _gcd_import
        File "<frozen importlib._bootstrap>", line 1354, in _find_and_load
        File "<frozen importlib._bootstrap>", line 1304, in _find_and_load_unlocked
        File "<frozen importlib._bootstrap>", line 488, in _call_with_frames_removed
        File "<frozen importlib._bootstrap>", line 1381, in _gcd_import
        File "<frozen importlib._bootstrap>", line 1354, in _find_and_load
        File "<frozen importlib._bootstrap>", line 1325, in _find_and_load_unlocked
        File "<frozen importlib._bootstrap>", line 929, in _load_unlocked
        File "<frozen importlib._bootstrap_external>", line 994, in exec_module
        File "<frozen importlib._bootstrap>", line 488, in _call_with_frames_removed
        File "/private/var/folders/1h/vpctqsm57x32h6_lcv8v2c0c0000gn/T/pip-build-env-rmmozt_b/overlay/lib/python3.12/site-packages/setuptools/__init__.py", line 16, in <module>
          import setuptools.version
        File "/private/var/folders/1h/vpctqsm57x32h6_lcv8v2c0c0000gn/T/pip-build-env-rmmozt_b/overlay/lib/python3.12/site-packages/setuptools/version.py", line 1, in <module>
          import pkg_resources
        File "/private/var/folders/1h/vpctqsm57x32h6_lcv8v2c0c0000gn/T/pip-build-env-rmmozt_b/overlay/lib/python3.12/site-packages/pkg_resources/__init__.py", line 2172, in <module>
          register_finder(pkgutil.ImpImporter, find_on_path)
                          ^^^^^^^^^^^^^^^^^^^
      AttributeError: module 'pkgutil' has no attribute 'ImpImporter'. Did you mean: 'zipimporter'?
      [end of output]
  
  note: This error originates from a subprocess, and is likely not a problem with pip.
error: subprocess-exited-with-error

× Getting requirements to build wheel did not run successfully.
│ exit code: 1
╰─> See above for output.

note: This error originates from a subprocess, and is likely not a problem with pip.

Issue resolved when reverted to python 3.10

PyKoopman for coupled PDEs

I would like to apply PyKoopman to a coupled PDE, let's say:

$\frac{\partial X}{\partial t} = \frac{\partial X}{\partial z} + T$
$\frac{\partial T}{\partial t} = X - \frac{\partial T}{\partial z}$

After discretizing for z and solving this system with some initial values I have the matrices t, z, X, T.
t.shape = (150,)
z.shape = (151,)
X.shape = (150, 151)
T.shape= (150, 151)

In which form can I input X and T into model.fit(x, dt=dt)? Does pyKoopman currently support this?

nndmd

Could you please inform me if your NNMD code is only suitable for very simple systems? When training with my dataset, I've found the results to be very poor, particularly at the sudden change points of some systems. The performance of the predict function is notably inadequate, as it seems capable only of fitting very smooth curves and struggles with instant transitions. Does this library have such an issue? I hope you can point out the parts of the code related to this problem so that I can modify it more effectively.

Installation error

I was just experimenting with this package. I realized that the installation doesn't work unless installed in the developer mode.

To be more specific the following command doesn't work:
pk,Koopman()

Issues in function _radial_basis_functions.py

This is the error message I got:

"pykoopman/observables/_radial_basis_functions.py:259
match self.rbf_type:
^
SyntaxError: invalid syntax"

Please help. Thanks

[enhancement] DeepKoopman regressor

Hello,
I took a quick look at your library and got surprised that no deepkoopman technique was implemented.
Do you plan to include it in the future ? ( Even if this may lead to include a deep-learning dependency to your project. )

I would be happy to try to submit this feature if you interested to include it.

Best regards

Fitting

Hello,
I am trying to rerun the tutorial code or get the observables names using the polynomial class but every time I get an error. When trying to obtain the get_features_names it isnt defined in the Base file and for the observable the model.koopman(obs=pk.observable.Polynomial(degree =2) regression= EDMDc()) always gives the error that model is not fitted. The tutorial I am trying to follow is the eigen functions with slow manifold and even running it on my pc doesn't work. Could anyone help me out?

dynamicslab / pykoopman Goto Github PK

pykoopman's People

Contributors

Stargazers

Watchers

Forkers

pykoopman's Issues

Recommend Projects

Recommend Topics

Recommend Org

Jobs