Getting started


Before trying out pyMOR, you need to install it. We provide packages for Ubuntu via our PPA:

sudo apt-add-repository ppa:pymor/stable
sudo apt-get update
sudo apt-get install python-pymor python-pymor-demos python-pymor-doc

Daily snapshots can be installed by using the pymor/daily PPA instead of pymor/stable. The current release can also be installed via pip. Please take a look at our README file for further details. The README also contains instructions for setting up a development environment for working on pyMOR itself.

Trying it out

While we consider pyMOR mainly as a library for building MOR applications, we ship a few example scripts. These can be found in the src/pymordemos directory of the source repository. Try launching one of them using the pymor-demo script contained in the python-pymor-demos package:

pymor-demo thermalblock --plot-err --plot-solutions 3 2 3 32

The demo scripts can also be launched directly from the source tree:

./ --plot-err --plot-solutions 3 2 3 32

This will solve and reduce the so called thermal block problem using the reduced basis method with a greedy basis generation algorithm. The thermal block problem consists in solving the stationary diffusion problem

- ∇ ⋅ [ d(x, μ) ∇ u(x, μ) ] = 1     for x in Ω
                  u(x, μ)   = 0     for x in ∂Ω

on the domain Ω = [0,1]^2 for the unknown u. The domain is partitioned into XBLOCKS x YBLOCKS blocks (XBLOCKS and YBLOCKS are the first two arguments to The thermal conductivity d(x, μ) is constant on each block (i,j) with value μ_ij:

|        |        |        |
|  μ_11  |  μ_12  |  μ_13  |
|        |        |        |
|        |        |        |
|  μ_21  |  μ_22  |  μ_23  |
|        |        |        |

The real numbers μ_ij form the XBLOCKS x YBLOCKS - dimensional parameter on which the solution depends.

Running will first produce plots of two detailed solutions of the problem for different randomly chosen parameters using linear finite elements. (The size of the grid can be controlled via the --grid parameter. The randomly chosen parameters will actually be the same for each run, since a the random generator is initialized with a fixed default seed in new_random_state.)

After closing the window, the reduced basis for model order reduction is generated using a greedy search algorithm with error estimator. The third parameter SNAPSHOTS of determines how many different values per parameter component μ_ij should be considered. I.e. the parameter training set for basis generation will have the size SNAPSHOTS^(XBLOCKS x YBLOCKS). After the basis of size 32 (the last parameter) has been computed, the quality of the obtained reduced model (on the 32-dimensional reduced basis space) is evaluated by comparing the solutions of the reduced and detailed models for new, randomly chosen parameters. Finally, plots of the detailed and reduced solutions, as well as the difference between the two, are displayed for the random parameter which maximises reduction error.

The thermalblock demo explained

In the following we will walk through the thermal block demo step by step in an interactive Python shell. We assume that you are familiar with the reduced basis method and that you know the basics of Python programming as well as working with NumPy. (Note that our code will differ a bit from as we will hardcode the various options the script offers and leave out some features.)

First, start a Python shell. We recommend using IPython


You can paste the following input lines starting with >>> by copying them to the system clipboard and then executing


inside the IPython shell.

First, we will import the most commonly used methods and classes of pyMOR by executing:

>>> from pymor.basic import *
Loading pymor version 0.3.0

Next we will instantiate a class describing the analytical problem we want so solve. In this case, a ThermalBlockProblem:

>>> p = ThermalBlockProblem(num_blocks=(3, 2))

We want to discretize this problem using the finite element method. We could do this by hand, creating a Grid, instatiating DiffusionOperatorP1 finite element diffusion operators for each subblock of the domain, forming a LincombOperator to represent the affine decomposition, instantiating a L2ProductFunctionalP1 as right hand side, and putting it all together into a StationaryDiscretization. However, since ThermalBlockProblem derives form EllipticProblem, we can use a predifined discretizer to do the work for us. In this case, we use discretize_elliptic_cg:

>>> d, d_data = discretize_elliptic_cg(p, diameter=1. / 100.)

d is the StationaryDiscretization which has been created for us, whereas d_data contains some additional data, in this case the Grid and the BoundaryInfo which have been created during discretization. We can have a look at the grid,

>>> print(d_data['grid'])
Tria-Grid on domain [0,1] x [0,1]
x0-intervals: 100, x1-intervals: 100
faces: 40000, edges: 60200, vertices: 20201

and, as always, we can display its class documentation using help(d_data['grid']), or in the case of IPython d_data['grid']?.

Let’s solve the thermal block problem and visualize the solution:

>>> U = d.solve([1.0, 0.1, 0.3, 0.1, 0.2, 1.0])
>>> d.visualize(U, title='Solution')
00:45|discretizations.basic.StationaryDiscretization: Solving ThermalBlock_CG for {diffusion: [1.0, 0.1, 0.3, 0.1, 0.2, 1.0]} ...

Each class in pyMOR that describes a Parameter dependent mathematical object, like the StationaryDiscretization in our case, derives from Parametric and determines the Parameters it expects during __init__ by calling build_parameter_type. The resulting ParameterType is stored in the object’s parameter_type attribute. Let us have a look:

>>> print(d.parameter_type)
{diffusion: (2, 3)}

This tells us, that the Parameter which ~pymor.discretizations.interfaces.DiscretizationInterface.solve expects should be a dictionary with one key 'diffusion' whose value is a NumPy array of shape (2, 3), corresponding to the block structure of the problem. However, by using the parse_parameter method, pyMOR is smart enough to correctly parse the input [1.0, 0.1, 0.3, 0.1, 0.2, 1.0].

Next we want to use the greedy algorithm to reduce the problem. For this we need to choose a basis extension algorithm as well as a reductor which will perform the actual RB-projection. We will use gram_schmidt_basis_extension and reduce_coercive. The latter will also assemble an error estimator to estimate the reduction error. This will significantly speed up the basis generation, as we will only need to solve the high-dimensional problem for those parameters in the training set which are actually selected for basis extension. To control the condition of the reduced system matrix, we must ensure that the generated basis is orthonormal w.r.t. the H1-product on the solution space. For this we pass the h1_product attribute of the discretization to the basis extension algorithm. We pass the same product to the reductor which uses it for computing the Riesz representatives required for error estimation. Moreover, we have to provide a ParameterFunctional which computes a lower bound for the coercivity of the problem for a given parameter.

>>> from functools import partial
>>> extension_algorithm = partial(gram_schmidt_basis_extension, product=d.h1_product)
>>> reductor = partial(
...     reduce_coercive,
...     error_product=d.h1_product,
...     coercivity_estimator=ExpressionParameterFunctional('min(diffusion)', d.parameter_type)
... )

Moreover, we need to select a Parameter training set. The discretization d already comes with a ParameterSpace which it has inherited from the analytical problem. We can sample our parameters from this space, which is a CubicParameterSpace. E.g.:

>>> samples = d.parameter_space.sample_uniformly(4)
>>> print(samples[0])
{diffusion: [0.1, 0.1, 0.1, 0.1, 0.1, 0.1]}

Now we start the basis generation:

>>> greedy_data = greedy(d, reductor, samples,
...                      extension_algorithm=extension_algorithm,
...                      use_estimator=True,
...                      max_extensions=32)
01:44 greedy: Started greedy search on 4096 samples
01:44 greedy: Reducing ...
01:44 |   reduce_coercive: RB projection ...
01:45 |   reduce_coercive: Assembling error estimator ...
01:45 |   |   reduce_residual: Estimating residual range ...
01:45 |   |   |   estimate_image_hierarchical: Estimating image for basis vector -1 ...
01:45 |   |   |   estimate_image_hierarchical: Orthonormalizing ...
01:45 |   |   reduce_residual: Projecting residual operator ...
01:45 greedy: Estimating errors ...
01:47 greedy: Maximum error after 0 extensions: 9.86736953629 (mu = {diffusion: [0.1, 0.1, 0.1, 0.1, 0.1, 0.1]})
01:47 greedy: Computing solution snapshot for mu = {diffusion: [0.1, 0.1, 0.1, 0.1, 0.1, 0.1]} ...
01:47 |   StationaryDiscretization: Solving ThermalBlock_CG for {diffusion: [0.1, 0.1, 0.1, 0.1, 0.1, 0.1]} ...
01:47 greedy: Extending basis with solution snapshot ...
01:47 greedy: Reducing ...
01:47 greedy: Estimating errors ...
01:50 greedy: Maximum error after 1 extensions: 3.30998133771 (mu = {diffusion: [0.1, 0.1, 1.0, 0.1, 0.1, 0.1]})
01:50 greedy: Computing solution snapshot for mu = {diffusion: [0.1, 0.1, 1.0, 0.1, 0.1, 0.1]} ...
01:50 |   StationaryDiscretization: Solving ThermalBlock_CG for {diffusion: [0.1, 0.1, 1.0, 0.1, 0.1, 0.1]} ...
01:50 greedy: Extending basis with solution snapshot ...
03:52 greedy: Maximum number of 32 extensions reached.
03:52 greedy: Reducing once more ...
03:55 greedy: Greedy search took 130.547789097 seconds

The max_extensions parameter defines how many basis vectors we want to obtain. greedy_data is a dictionary containing various data that has been generated during the run of the algorithm:

>>> print(greedy_data.keys())
['reduction_data', 'reconstructor', 'time', 'basis', 'extensions', 'reduced_discretization', 'max_errs', 'max_err_mus']

The most important items are 'reduced_discretization' and 'reconstructor' which hold the reduced Discretization obtained from applying our reductor with the final reduced basis, as well as a reconstructor to reconstruct detailed solutions from the reduced solution vectors. The reduced basis is stored as 'basis' item.

>>> rd = greedy_data['reduced_discretization']
>>> rc = greedy_data['reconstructor']
>>> rb = greedy_data['basis']

All vectors in pyMOR are stored in so called VectorArrays. For example the solution U computed above is given as a VectorArray of length 1. For the reduced basis we have:

>>> print(type(rb))
<class 'pymor.vectorarrays.numpy.NumpyVectorArray'>
>>> print(len(rb))
>>> print(rb.dim)

Let us check if the reduced basis really is orthonormal with respect to the H1-product. For this we use the apply2 method:

>>> import numpy as np
>>> gram_matrix = d.h1_product.apply2(rb, rb)
>>> print(np.max(np.abs(gram_matrix - np.eye(32))))

Looks good! We can now solve the reduced model for the same parameter as above. The result is a vector of coefficients w.r.t. the reduced basis, which is currently stored in rb. To form the linear combination, we can use the reconstructor:

>>> u = rd.solve([1.0, 0.1, 0.3, 0.1, 0.2, 1.0])
>>> print(u)
[[  5.79477471e-01   5.91289054e-02   1.89924036e-01   1.89149529e-02
    1.81103127e-01   2.69920752e-02  -1.79611519e-01   7.99676272e-03
    1.54092560e-01   5.76326362e-02   1.97982347e-01  -2.13775254e-02
    3.12892660e-02  -1.27037440e-01  -1.51352508e-02   3.36101087e-02
    2.05779889e-02  -4.96445984e-03   3.21176662e-02  -2.52674851e-02
    2.92150040e-02   3.23570362e-03  -4.14288199e-03   5.48325425e-03
    4.10728945e-03   1.59251955e-03  -9.23470903e-03  -2.57483574e-03
   -2.52451212e-03  -5.08125873e-04   2.71427033e-03   5.83210112e-05]]
>>> U_red = rc.reconstruct(u)
>>> print(U_red.dim)

Finally we compute the reduction error and display the reduced solution along with the detailed solution and the error:

>>> ERR = U - U_red
>>> print(d.h1_norm(ERR))
[ 0.00944595]
>>> d.visualize((U, U_red, ERR),
...             legend=('Detailed', 'Reduced', 'Error'),
...             separate_colorbars=True)

We can nicely observe that, as expected, the error is maximized along the jumps of the diffusion coeffient.

Learning more

As a next step, you should read our Technical Overview which discusses the most important concepts and design decisions behind pyMOR. After that you should be fit to delve into the reference documentation.

Should you have any problems regarding pyMOR, questions or feature requests, do not hestitate to contact us at our mailing list!