Description¶
This package provides engineering-related classes and functions, including:
A waveform class that is a first-class object. For example:
>>> import copy, numpy, peng >>> obj_a=peng.Waveform( ... indep_vector=numpy.array([1, 2, 3]), ... dep_vector=numpy.array([10, 20, 30]), ... dep_name='obj_a' ... ) >>> obj_b = obj_a*2 >>> print(obj_b) Waveform: obj_a*2 Independent variable: [ 1, 2, 3 ] Dependent variable: [ 20, 40, 60 ] Independent variable scale: LINEAR Dependent variable scale: LINEAR Independent variable units: (None) Dependent variable units: (None) Interpolating function: CONTINUOUS >>> obj_c = copy.copy(obj_b) >>> obj_a == obj_b False >>> obj_b == obj_c True
Numerous functions are provided (trigonometric, calculus, transforms, etc.) and creating new functions that operate on waveforms is simple since all of their relevant information can be accessed through properties
Handling numbers represented in engineering notation, obtaining their constituent components and converting to and from regular floats. For example:
>>> import peng >>> x = peng.peng(1346, 2, True) >>> x ' 1.35k' >>> peng.peng_float(x) 1350.0 >>> peng.peng_int(x) 1 >>> peng.peng_frac(x) 35 >>> str(peng.peng_mant(x)) '1.35' >>> peng.peng_power(x) EngPower(suffix='k', exp=1000.0) >>> peng.peng_suffix(x) 'k'
Pretty printing Numpy vectors. For example:
>>> from __future__ import print_function >>> import peng >>> header = 'Vector: ' >>> data = [1e-3, 20e-6, 30e+6, 4e-12, 5.25e3, -6e-9, 70, 8, 9] >>> print( ... header+peng.pprint_vector( ... data, ... width=30, ... eng=True, ... frac_length=1, ... limit=True, ... indent=len(header) ... ) ... ) Vector: [ 1.0m, 20.0u, 30.0M, ... 70.0 , 8.0 , 9.0 ]
Formatting numbers represented in scientific notation with a greater degree of control and options than standard Python string formatting. For example:
>>> import peng >>> peng.to_scientific_string( ... number=99.999, ... frac_length=1, ... exp_length=2, ... sign_always=True ... ) '+1.0E+02'
Interpreter¶
The package has been developed and tested with Python 2.6, 2.7, 3.3, 3.4 and 3.5 under Linux (Debian, Ubuntu), Apple OS X and Microsoft Windows
Installing¶
$ pip install peng
Documentation¶
Available at Read the Docs
Contributing¶
Abide by the adopted code of conduct
Fork the repository from GitHub and then clone personal copy [1]:
$ git clone \ https://github.com/[github-user-name]/peng.git Cloning into 'peng'... ... $ cd peng $ export PENG_DIR=${PWD}
Install the project’s Git hooks and build the documentation. The pre-commit hook does some minor consistency checks, namely trailing whitespace and PEP8 compliance via Pylint. Assuming the directory to which the repository was cloned is in the
$PENG_DIR
shell environment variable:$ ${PENG_DIR}/sbin/complete-cloning.sh Installing Git hooks Building peng package documentation ...
Ensure that the Python interpreter can find the package modules (update the
$PYTHONPATH
environment variable, or use sys.paths(), etc.)$ export PYTHONPATH=${PYTHONPATH}:${PENG_DIR}
Install the dependencies (if needed, done automatically by pip):
- Astroid (Python 2.6: older than 1.4, Python 2.7 or newer: 1.3.8 or newer)
- Cog (2.4 or newer)
- Coverage (3.7.1 or newer)
- Decorator (3.4.2 or newer)
- Docutils (0.12 or newer)
- Funcsigs (Python 2.x only, 0.4 or newer)
- Inline Syntax Highlight Sphinx Extension (0.2 or newer)
- Mock (Python 2.x only, 1.0.1 or newer)
- Nose (Python 2.6: 1.0.0 or newer)
- Numpy (1.8.2 or newer)
- Pexdoc (1.0.0 or newer)
- Pmisc (1.0.0 or newer)
- Py.test (2.7.0 or newer)
- PyParsing (2.0.7 or newer)
- Pylint (Python 2.6: 1.3 or newer and older than 1.4, Python 2.7 or newer: 1.3.1 or newer)
- Pytest-coverage (1.8.0 or newer)
- Pytest-xdist (optional, 1.8.0 or newer)
- ReadTheDocs Sphinx theme (0.1.9 or newer)
- Scipy (0.13.3 or newer)
- Six (1.4.0 or newer)
- Sphinx (1.2.3 or newer)
- Tox (1.9.0 or newer)
- Virtualenv (13.1.2 or newer)
Implement a new feature or fix a bug
Write a unit test which shows that the contributed code works as expected. Run the package tests to ensure that the bug fix or new feature does not have adverse side effects. If possible achieve 100% code and branch coverage of the contribution. Thorough package validation can be done via Tox and Py.test:
$ tox GLOB sdist-make: .../peng/setup.py py26-pkg inst-nodeps: .../peng/.tox/dist/peng-...zip
Setuptools can also be used (Tox is configured as its virtual environment manager) [2]:
$ python setup.py tests running tests running egg_info writing requirements to peng.egg-info/requires.txt writing peng.egg-info/PKG-INFO ...
Tox (or Setuptools via Tox) runs with the following default environments:
py26-pkg
,py27-pkg
,py33-pkg
,py34-pkg
andpy35-pkg
[3]. These use the Python 2.6, 2.7, 3.3, 3.4 and 3.5 interpreters, respectively, to test all code in the documentation (both in Sphinx*.rst
source files and in docstrings), run all unit tests, measure test coverage and re-build the exceptions documentation. To pass arguments to Py.test (the test runner) use a double dash (--
) after all the Tox arguments, for example:$ tox -e py27-pkg -- -n 4 GLOB sdist-make: .../peng/setup.py py27-pkg inst-nodeps: .../peng/.tox/dist/peng-...zip ...
Or use the
-a
Setuptools optional argument followed by a quoted string with the arguments for Py.test. For example:$ python setup.py tests -a "-e py27-pkg -- -n 4" running tests ...
There are other convenience environments defined for Tox [4]:
py26-repl
,py27-repl
,py33-repl
,py34-repl
andpy35-repl
run the Python 2.6, 2.7, 3.3, 3.4 or 3.5 REPL, respectively, in the appropriate virtual environment. Thepeng
package is pip-installed by Tox when the environments are created. Arguments to the interpreter can be passed in the command line after a double dash (--
)py26-test
,py27-test
,py33-test
,py34-test
andpy35-test
run py.test using the Python 2.6, 2.7, 3.3, 3.4 or Python 3.5 interpreter, respectively, in the appropriate virtual environment. Arguments to py.test can be passed in the command line after a double dash (--
) , for example:$ tox -e py34-test -- -x test_eng.py GLOB sdist-make: [...]/peng/setup.py py34-test inst-nodeps: [...]/peng/.tox/dist/peng-[...].zip py34-test runtests: PYTHONHASHSEED='680528711' py34-test runtests: commands[0] | [...]py.test -x test_eng.py ==================== test session starts ==================== platform linux -- Python 3.4.2 -- py-1.4.30 -- [...] ...
py26-cov
,py27-cov
,py33-cov
,py34-cov
andpy35-cov
test code and branch coverage using the Python 2.6, 2.7, 3.3, 3.4 or 3.5 interpreter, respectively, in the appropriate virtual environment. Arguments to py.test can be passed in the command line after a double dash (--
). The report can be found in${PENG_DIR}/.tox/py[PV]/usr/share/peng/tests/htmlcov/index.html
where[PV]
stands for26
,27
,33
,34
or35
depending on the interpreter used
Verify that continuous integration tests pass. The package has continuous integration configured for Linux (via Travis) and for Microsoft Windows (via Appveyor). Aggregation/cloud code coverage is configured via Codecov. It is assumed that the Codecov repository upload token in the Travis build is stored in the
${CODECOV_TOKEN}
environment variable (securely defined in the Travis repository settings page). Travis build artifacts can be transferred to Dropbox using the Dropbox Uploader script (included for convenience in the${PENG_DIR}/sbin
directory). For an automatic transfer that does not require manual entering of authentication credentials place the APPKEY, APPSECRET, ACCESS_LEVEL, OAUTH_ACCESS_TOKEN and OAUTH_ACCESS_TOKEN_SECRET values required by Dropbox Uploader in the in the${DBU_APPKEY}
,${DBU_APPSECRET}
,${DBU_ACCESS_LEVEL}
,${DBU_OAUTH_ACCESS_TOKEN}
and${DBU_OAUTH_ACCESS_TOKEN_SECRET}
environment variables, respectively (also securely defined in Travis repository settings page)Document the new feature or bug fix (if needed). The script
${PENG_DIR}/sbin/build_docs.py
re-builds the whole package documentation (re-generates images, cogs source files, etc.):$ ${PUTIL_DIR}/sbin/build_docs.py -h usage: build_docs.py [-h] [-d DIRECTORY] [-r] [-n NUM_CPUS] [-t] Build peng package documentation optional arguments: -h, --help show this help message and exit -d DIRECTORY, --directory DIRECTORY specify source file directory (default ../peng) -r, --rebuild rebuild exceptions documentation. If no module name is given all modules with auto-generated exceptions documentation are rebuilt -n NUM_CPUS, --num-cpus NUM_CPUS number of CPUs to use (default: 1) -t, --test diff original and rebuilt file(s) (exit code 0 indicates file(s) are identical, exit code 1 indicates file(s) are different)
Output of shell commands can be automatically included in reStructuredText source files with the help of Cog and the
docs.support.term_echo
module.-
docs.support.term_echo.
ste
(command, nindent, mdir, fpointer) Simplified terminal echo; prints STDOUT resulting from a given Bash shell command (relative to the package
sbin
directory) formatted in reStructuredTextParameters: - command (string) – Bash shell command, relative to
${PUTIL_DIR}/sbin
- nindent (integer) – Indentation level
- mdir (string) – Module directory
- fpointer (function object) – Output function pointer. Normally is
cog.out
butprint
or other functions can be used for debugging
For example:
.. This is a reStructuredText file snippet .. [[[cog .. import os, sys .. from docs.support.term_echo import term_echo .. file_name = sys.modules['docs.support.term_echo'].__file__ .. mdir = os.path.realpath( .. os.path.dirname( .. os.path.dirname(os.path.dirname(file_name)) .. ) .. ) .. [[[cog ste('build_docs.py -h', 0, mdir, cog.out) ]]] .. code-block:: bash $ ${PUTIL_DIR}/sbin/build_docs.py -h usage: build_docs.py [-h] [-d DIRECTORY] [-r] [-n NUM_CPUS] [-t] [module_name [module_name ...]] ... .. ]]]
- command (string) – Bash shell command, relative to
-
docs.support.term_echo.
term_echo
(command, nindent=0, env=None, fpointer=None, cols=60) Terminal echo; prints STDOUT resulting from a given Bash shell command formatted in reStructuredText
Parameters: - command (string) – Bash shell command
- nindent (integer) – Indentation level
- env (dictionary) – Environment variable replacement dictionary. The Bash
command is pre-processed and any environment variable
represented in the full notation (
${...}
) is replaced. The dictionary key is the environment variable name and the dictionary value is the replacement value. For example, if command is'${PYTHON_CMD} -m "x=5"'
and env is{'PYTHON_CMD':'python3'}
the actual command issued is'python3 -m "x=5"'
- fpointer (function object) – Output function pointer. Normally is
cog.out
butprint
or other functions can be used for debugging - cols (integer) – Number of columns of output
Similarly Python files can be included in docstrings with the help of Cog and the
docs.support.incfile
module-
docs.support.incfile.
incfile
(fname, fpointer, lrange='1, 6-', sdir=None) Includes a Python source file in a docstring formatted in reStructuredText
Parameters: - fname (string) – File name, relative to environment variable
${TRACER_DIR}
- fpointer (function object) – Output function pointer. Normally is
cog.out
butprint
or other functions can be used for debugging - lrange (string) – Line range to include, similar to Sphinx literalinclude directive
- sdir (string) – Source file directory. If None the
${TRACER_DIR}
environment variable is used if it is defined, otherwise the directory where thedocs.support.incfile
module is located is used
For example:
def func(): """ This is a docstring. This file shows how to use it: .. =[=cog .. import docs.support.incfile .. docs.support.incfile.incfile('func_example.py', cog.out) .. =]= .. code-block:: python # func_example.py if __name__ == '__main__': func() .. =[=end=]= """ return 'This is func output'
- fname (string) – File name, relative to environment variable
-
Footnotes
[1] | All examples are for the bash shell |
[2] | It appears that Scipy dependencies do not include Numpy (as they should) so running the tests via Setuptools will typically result in an error. The peng requirement file specifies Numpy before Scipy and this installation order is honored by Tox so running the tests via Tox sidesteps Scipy’s broken dependency problem but requires Tox to be installed before running the tests (Setuptools installs Tox if needed) |
[3] | It is assumed that all the Python interpreters are in the executables path. Source code for the interpreters can be downloaded from Python’s main site |
[4] | Tox configuration largely inspired by Ionel’s codelog |
Changelog¶
- 0.9.11 [2016-04-15]:
- Created new APIs in the exh module to simplify adding and conditionally raising exceptions that can be auto-documented with the exdoc module
- Homogenized API arguments in several pcsv module functions
- Bug fixes
- Documentation updates
- 0.9.10 [2016-03-10]: Final release of 0.9.10 branch
- 0.9.10rc1 [2016-03-09]: Apple OS X compatibility changes. Reduced memory consumption during exception auto-documentation process. Bug fixes
- 0.9.9 [2016-01-27]: Fixed documentation bugs that were causing errors with Sphinx 1.3.5+
- 0.9.8 [2016-01-22]: Bug fixes
- 0.9.7 [2016-01-22]: Enhanced control of exceptions automatic documentation output
- 0.9.6 [2016-01-20]: Bug fixes
- 0.9.5 [2016-01-08]: Bug fixes
- 0.9.4 [2015-12-18]: Minor documentation update regarding continuous integration setup
- 0.9.3 [2015-12-17]: Fixed critical bug in plot module that prevented figures without any axis labels from being generated
- 0.9.2 [2015-12-15]: Speed improvements in plot module
- 0.9.1 [2015-12-01]: Final release of 0.9.1 branch
- 0.9.1rc5 [2015-12-01]: Fixed documentation URL in top-level README.rst
- 0.9.1rc4 [2015-12-01]: Fixed bug in top-level README.rst verification
- 0.9.1rc3 [2015-12-01]:
- Documentation updates
- Package verification improvements
- 0.9.1rc2 [2015-12-01]: Fixed top-level README.rst file
- 0.9.1rc1 [2015-11-30]: Initial public release
License¶
The MIT License (MIT)
Copyright (c) 2013-2016 Pablo Acosta-Serafini
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the “Software”), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.