Repository files navigation

• Examples using thenAG Library for Python
• How to install thenAG Library for Python
• How to run the Jupyter notebook examples
• List of Chapters in thenAG Library for Python
• Useful links

This repository contains examples and demonstrations using thenAG Library for Python. ThenAG Library for Python contains 1900+ functions spanning many areas of numerical computing and data science.

Designed to work alongside the open source Python packages,Numpy andScipy, ThenAG Library for Python can augment your computational workflow in many areas.

• Correlation and regression analysis
• Curve and surface fitting
• Dimensionality Reduction
• Global Optimization
• Linear Algebra
• Local optimization
• Multivariate methods
• Nearest Correlation Matrices
• Operations Research
• Random number generation
• Roots
• Time series analysis

In addition to those presented here, ThenAG Library for Python ships with a set of usage examples. To see them all, run the following command

python -m naginterfaces.library.examples --locate

In this section we illustrate how to install thenAG Library for Python, request a Trial Licence and make sure the Library is working. Details and further information regarding the installation can be foundhere.

Note Before starting make sure you have access to a host that has Python 3 (3.4 or more recent).

Installing thenAG Library is done using thepip package manager, fire-up a terminal and create a Python 3 virtual environment where to install and test thenAG Library

guest@nag-37:~$ python3 -m venv nag3guest@nag-37:~$ . nag3/bin/activate(nag3) guest@nag-37:~$

Now usepip to install thenAG Library for Python

(nag3) guest@nag-37:~$ python -m pip install --extra-index-url https://www.nag.com/downloads/py/naginterfaces_nag naginterfaces

or if you prefer the version of the package that relies on Intel MKL for optimized linear algebra routines, then use

(nag3) guest@nag-37:~$ python -m pip install --extra-index-url https://www.nag.com/downloads/py/naginterfaces_mkl naginterfaces

The output should be similar to

Collecting naginterfaces  Downloading https://www.nag.com/downloads/py/naginterfaces_nag/naginterfaces/naginterfaces-27.1.0.0-py2.py3-none-linux_x86_64.whl (55.8MB)    100% |████████████████████████████████| 55.8MB 21kB/s Collecting numpy>=1.15 (from naginterfaces)  Downloading https://files.pythonhosted.org/packages/45/b2/6c7545bb7a38754d63048c7696804a0d947328125d81bf12beaa692c3ae3/numpy-1.19.5-cp36-cp36m-manylinux1_x86_64.whl (13.4MB)    100% |████████████████████████████████| 13.4MB 70kB/s Installing collected packages: numpy, naginterfacesSuccessfully installed naginterfaces-27.1.0.0 numpy-1.19.5

The output indicates that the installation was successful.

The next step is to get the licensing info (product code andKUSARI ID) and use it to request a licence. From the same virtual terminal, try

(nag3) guest@nag-37:~$ python -m naginterfaces.kusari

The output should be similar to

The *n*AG Library for Python on this platform usesunderlying Library NLL6I271VL.This Library has been installed as part of the packageand it requires a valid licence key.No such key could be validated:the key may not have been installed correctly orit may have expired.The Kusari licence-check utility reports the following:User: guestDirectory: /home/guestNAG_KUSARI_FILE=""File /home/guest/nag.key does not exist-------------------------------------------------------------------------------Error: Licence not found; this product requires a key for NLL6I271VLThe above information has been generated on machine nag-37For information on how to obtain a licence, please seehttps://www.nag.com/numeric/py/nagdoc_latest/naginterfaces.kusari.htmlKUSARI ID = "ADLXt-adEclJLmvnxlrU2sseteZoo,RopA-Ld"

Thetwo important bits are the

1. product code shown asunderlying Library NLL6I271VL. which identifies the licence to request, and

2. KUSARI ID shown asKUSARI ID = "ADLXt-adEclJLmvnxlrU2sseteZoo,RopA-Ld" which identifies the host you are running the library on.

Note that theproduct code andKUSARI ID can be different from the previous example.

With these, you are set tocontactnAG and request a trial licence.

The trial licence is a plain text chunk similar to

NLL6I271V TRIAL 2021/01/27 "RverXn0Pc-Ib?ctdgF=Wpis2j7I"

Save or copy the text into the file/home/guest/nag.key.

The final step is to make sure the licence is valid and the library is working as expected.

The last step is to make sure the licence was correctly stored and that thenAG Library is working correctly. From the same virtual terminal re-run the Kusari licence module

(nag3) guest@nag-37:~$ python -m naginterfaces.kusari

This time the output should be similar to

Licence available; the required NLL6I271VL licence key for this product is validTRIAL licence, 27 days remaining (licence from file)

Now let's try a more interesting example (list of optimization examples)

This command runs the example for theFOAS (First-Order Active set method) solver and minimizes the Rosenbrock 2D function.

(nag3) guest@nag-37:~$ python -m naginterfaces.library.examples.opt.handle_solve_bounds_foas_ex

Should generate an outputsimilar to

Trying:    main()Expecting:    naginterfaces.library.opt.handle_solve_bounds_foas Python Example Results.    Minimizing a bound-constrained Rosenbrock problem.     E04KF, First order method for bound-constrained problems...     Status: converged, an optimal solution was found     Value of the objective             4.00000E-02    ...ok

indicating that the example was successfully executed. The source code can be foundhere.

To display the full list of example source files on disk, but not run them, execute

python -m naginterfaces.library.examples --locate

All examples may be executed sequentially by running

python -m naginterfaces.library.examples

Runpython -m naginterfaces.library.examples --help to see any additional usage.

This section briefly illustrates how to setup a host in order to open and run theJupyter notebooks provided in this repository.Before running the notebooks make sure thenAG Library is installed and working. Before starting, it is advised to readJupyter's installation page.

To install Jupyter, launch a terminal and activate the virtual environment used to install thenAG Library for Python

guest@nag-37:~$ . nag3/bin/activate(nag3) guest@nag-37:~$ pip install notebook matplotlibCollecting notebook  Downloading https://files.pythonhosted.org/packages/74/19/50cd38acf22e33370d01fef764355f1e3517f6e12b4fceb8d434ece4f8fd/notebook-6.2.0-py3-none-any.whl (9.5MB)    100% |████████████████████████████████| 9.5MB 115kB/s Collecting argon2-cffi (from notebook)...Successfully installed jupyter-client-6.1.11 jupyterlab-pygments-0.1.2 ... wcwidth-0.2.5

This indicates that Jupyter and matplotlib were successfully installed. The next section shows how to start the notebok interface and open an example.

To run an example, grab a copy of the notebook of interest and start up the notebook interface.For example, download theRosenbrock 2D optimization example notebookrosenbrock2d.ipynb into the current directory

(nag3) guest@nag-37:~$ curl -O https://raw.githubusercontent.com/numericalalgorithmsgroup/NAGPythonExamples/master/local_optimization/FOAS/rosenbrock2d.ipynb  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current                                 Dload  Upload   Total   Spent    Left  Speed100 61961  100 61961    0     0   382k      0 --:--:-- --:--:-- --:--:--  382k

and now open it usingjupyter-notebook

(nag3) guest@nag-37:~$ jupyter-notebook rosenbrock2d.ipynb[I 12:24:07.336 NotebookApp] Serving notebooks from local directory: /home/guest[I 12:24:07.336 NotebookApp] Jupyter Notebook 6.2.0 is running at:[I 12:24:07.336 NotebookApp] http://localhost:8888/?token=f1836a06799a92f25ef9966439bf3491b2f0960dcb51806d...

This command will fire-up your web browser and open therosenbrock2d.ipynb notebook, the window should be similar to

The following links take you to the relevant section in the official documentation

• library.anova - Analysis of Variance
• library.blas - Linear Algebra Support Routines
• library.blast - Further Linear Algebra Support Routines
• library.blgm - Linear Model Specification
• library.complex - Complex Arithmetic
• library.contab - Contingency Table Analysis
• library.correg - Correlation and Regression Analysis
• library.det - Determinants
• library.dot - Inner Products
• library.eigen - Eigenvalues and Eigenvectors
• library.fit - Curve and Surface Fitting
• library.glopt - Global Optimization of a Function
• library.ieee - IEEE Arithmetic
• library.info - Library Identification
• library.inteq - Integral Equations
• library.interp - Interpolation
• library.lapackeig - Least Squares and Eigenvalue Problems
• library.lapacklin - Linear Equations
• library.linsys - Simultaneous Linear Equations
• library.machine - Machine Constants
• library.math - Mathematical Constants
• library.matop - Matrix Operations, Including Inversion
• library.mesh - Mesh Generation
• library.mip - Operations Research
• library.mv - Multivariate Methods
• library.nonpar - Nonparametric Statistics
• library.numdiff - Numerical Differentiation
• library.ode - Ordinary Differential Equations
• library.omp - OpenMP Utilities
• library.opt - Minimizing or Maximizing a Function
• library.orthog - Orthogonalization
• library.pde - Partial Differential Equations
• library.quad - Quadrature
• library.rand - Random Number Generators
• library.rnla - Randomized Numerical Linear Algebra
• library.roots - Roots of One or More Transcendental Equations
• library.smooth - Smoothing in Statistics
• library.sort - Sorting and Searching
• library.sparse - Large Scale Linear Systems
• library.sparseig - Large Scale Eigenproblems
• library.specfun - Approximations of Special Functions
• library.stat - Simple Calculations on Statistical Data
• library.sum - Summation of Series
• library.surviv - Survival Analysis
• library.tsa - Time Series Analysis
• library.univar - Univariate Estimation
• library.wav - Wavelet Transforms
• library.zeros - Zeros of Polynomials

• nAG Library for Python Documentation
• Kusari licence module Documentation

Back to Top