create or adapt or include a units package #3852

jasongrout · 2008-08-14T17:47:02Z

See this sage-wiki page for some ideas.

Here are three possibilities:

ScientificPython's PhysicalQuantities package: http://dirac.cnrs-orleans.fr/ScientificPython/ScientificPythonManual/Scientific.Physics.PhysicalQuantities-module.html

It looks like this package is part of an actively maintained library, ScientificPython (http://dirac.cnrs-orleans.fr/plone/software/scientificpython/). Here is a link for a printing package for this module: http://python.net/crew/bhoel/PyLaTeX-0.2/html/PyLaTeX.UnitPrint.html

Unum: http://home.scarlet.be/be052320/Unum.html

It looks like this package is dead; last release was in 2005. It is GPL.

The Pyre units package (http://www.cacr.caltech.edu/projects/pyre/). This was last updated in 2005, I think (you have to check out pythia 0.8).
The enthought units package (which is apparently based on the pyre package), as well as the blockcanvas unit functions.

A thread about adding a units package to scipy: http://projects.scipy.org/pipermail/scipy-user/2005-March/004263.html

Another thread: http://aspn.activestate.com/ASPN/Mail/Message/scipy-user/3176352

Another page in a book mentioning two packages: http://books.google.com/books?id=j7QbD83-h8AC&pg=PA157&lpg=PA157&dq=python+physicalquantities&source=web&ots=C-JZyCFWn2&sig=x62bQ0jWyQerBGAvgJOBQuMQeWM&hl=en&sa=X&oi=book_result&resnum=10&ct=result

A recent (July 2008!) trac ticket about the enthought units package: https://svn.enthought.com/enthought/ticket/1524 Related checkins seem to include: https://svn.enthought.com/enthought/changeset/21093

Apparently the enthought units code is undergoing refactorization. See https://mail.enthought.com/pipermail/enthought-dev/2008-July/015717.html

I'd say we ought to keep our eye (or help out with) the enthought package refactorization.

Component: basic arithmetic

Author: David Ackerman, William Stein

Reviewer: Dan Drake

Merged: sage-4.4.alpha0

Issue created by migration from https://trac.sagemath.org/ticket/3852

jasongrout · 2008-08-14T17:53:19Z

comment:1

Okay, so that was four projects!

jasongrout · 2008-08-20T06:03:25Z

comment:3

yet another project recently started:

http://sourceforge.net/mailarchive/forum.php?thread_name=200808051930.57498.dsdale24%40gmail.com&forum_name=matplotlib-devel

williamstein · 2009-10-04T18:54:58Z

comment:5

Attachment: trac_3852.patch.gz

FIRST REFEREE REPORT on trac_3852.patch:

The doctest coverage is still very bad (38%):

wstein@sage:~/build/sage-4.1.1$ ./sage -coverage devel/sage/sage/symbolic/units.py 
----------------------------------------------------------------------
devel/sage/sage/symbolic/units.py
ERROR: Please define a s == loads(dumps(s)) doctest.
SCORE devel/sage/sage/symbolic/units.py: 38% (5 of 13)

Missing documentation:
         * evalunitdict():
         * unit_derivations_expr(v):
         * _sage_doc_(self):
         * str_to_unit(name):
         * __init__(self, data):
         * trait_names(self):
         * __getattr__(self, name):


Missing doctests:
         * vars_in_str(s):


Possibly wrong (function name doesn't occur in doctests):
         * convert_temperature(expr, target):

----------------------------------------------------------------------

The new code fails numerous doctests:

        sage -t  devel/sage/sage/symbolic/units.py # 4 doctests failed
        sage -t  devel/sage/sage/symbolic/expression.pyx # 2 doctests failed

The actual failures:


**********************************************************************
File "/scratch/wstein/build/sage-4.1.1/devel/sage/sage/symbolic/expression.pyx", line 5399:
    sage: (units.pressure.pascal*units.si_prefixes.kilo).convert(units.pressure.pounds_per_square_inch)
Exception raised:
    Traceback (most recent call last):
      File "/scratch/wstein/build/sage-4.1.1/local/bin/ncadoctest.py", line 1231, in run_one_test
        self.run_one_example(test, example, filename, compileflags)
      File "/scratch/wstein/build/sage-4.1.1/local/bin/sagedoctest.py", line 38, in run_one_example
        OrigDocTestRunner.run_one_example(self, test, example, filename, compileflags)
      File "/scratch/wstein/build/sage-4.1.1/local/bin/ncadoctest.py", line 1172, in run_one_example
        compileflags, 1) in test.globs
      File "<doctest __main__.example_137[10]>", line 1, in <module>
        (units.pressure.pascal*units.si_prefixes.kilo).convert(units.pressure.pounds_per_square_inch)###line 5399:
    sage: (units.pressure.pascal*units.si_prefixes.kilo).convert(units.pressure.pounds_per_square_inch)
      File "expression.pyx", line 5424, in sage.symbolic.expression.Expression.convert (sage/symbolic/expression.cpp:21250
)
      File "/scratch/wstein/build/sage-4.1.1/local/lib/python/site-packages/sage/symbolic/units.py", line 1064, in convert
        expr = expr.subs(z)
      File "expression.pyx", line 2777, in sage.symbolic.expression.Expression.substitute (sage/symbolic/expression.cpp:13
700)
      File "expression.pyx", line 1517, in sage.symbolic.expression.Expression.coerce_in (sage/symbolic/expression.cpp:959
1)
      File "parent_old.pyx", line 331, in sage.structure.parent_old.Parent._coerce_ (sage/structure/parent_old.c:4673)
      File "parent.pyx", line 429, in sage.structure.parent.Parent.coerce (sage/structure/parent.c:4806)
    TypeError: no canonical coercion from <type 'str'> to Symbolic Ring

...

File "/scratch/wstein/build/sage-4.1.1/devel/sage/sage/symbolic/units.py", line 31:
    e: units.force.dyne?
Exception raised:
    Traceback (most recent call last):
      File "/scratch/wstein/build/sage-4.1.1/local/bin/ncadoctest.py", line 1231, in run_one_test
        self.run_one_example(test, example, filename, compileflags)
      File "/scratch/wstein/build/sage-4.1.1/local/bin/sagedoctest.py", line 38, in run_one_example
        OrigDocTestRunner.run_one_example(self, test, example, filename, compileflags)
      File "/scratch/wstein/build/sage-4.1.1/local/bin/ncadoctest.py", line 1172, in run_one_example
        compileflags, 1) in test.globs
      File "<doctest __main__.example_0[6]>", line 1
         units.force.dyne?###line 31:
    e: units.force.dyne?
                         ^
     SyntaxError: invalid syntax
**********************************************************************
File "/scratch/wstein/build/sage-4.1.1/devel/sage/sage/symbolic/units.py", line 52:

**********************************************************************
File "/scratch/wstein/build/sage-4.1.1/devel/sage/sage/symbolic/units.py", line 52:
    e: t.convert(unit.charge.coulomb)
Expected:
    Traceback (most recent call last):
    ...
    ValueError: Incompatible units
Got:
    Traceback (most recent call last):
      File "/scratch/wstein/build/sage-4.1.1/local/bin/ncadoctest.py", line 1231, in run_one_test
        self.run_one_example(test, example, filename, compileflags)
      File "/scratch/wstein/build/sage-4.1.1/local/bin/sagedoctest.py", line 38, in run_one_example
        OrigDocTestRunner.run_one_example(self, test, example, filename, compileflags)
      File "/scratch/wstein/build/sage-4.1.1/local/bin/ncadoctest.py", line 1172, in run_one_example
        compileflags, 1) in test.globs
      File "<doctest __main__.example_0[11]>", line 1, in <module>
        t.convert(unit.charge.coulomb)###line 52:
    e: t.convert(unit.charge.coulomb)
    NameError: name 'unit' is not defined


...

sagetrac-edrex · 2009-11-04T21:22:31Z

comment:6

Google also turns up this existing module: http://www.inference.phy.cam.ac.uk/dimpy/

sagetrac-edrex · 2009-11-04T21:29:43Z

comment:7

dimpy helped me with my e&m hw. positive review :)

jasongrout · 2009-11-04T21:44:45Z

comment:8

Do you mean that tongue-in-cheek, or do you mean that you've gone through and made sure all of the objections above were taken care of?

sagetrac-edrex · 2009-11-05T03:32:15Z

comment:9

my apologies, I hadn't properly read the issue, much less the patch, and I meant to retract the post.

I was actually referring to the dimpy module, not the patch above, and it was definitely tongue-in-cheek. Upon further investigation I found that dimpy doesn't integrate well with sage's types system.

I've now read the patch and will try to find time to run the tests.

williamstein · 2010-01-21T02:54:02Z

Attachment: trac_3852-part2.patch.gz

sagetrac-schymans · 2010-01-26T15:26:44Z

comment:13

Attachment: trac_3852-part3.patch.gz

Hm, it is not easy to find out what units() is actually supposed to be used for..
In the notebook, units? gives:

File: /Users/sschym/Programs/sage/local/lib/python2.6/site-packages/sage/symbolic/units.py

Type: <type ‘instance’>

Definition: units( [noargspec] )

Docstring:

A collection of units of a some type.

    EXAMPLES:

    sage: units.power
    Collection power of units: cheval_vapeur horsepower watt

I could not find anything useful in the documentation, either. Do I have to compile the documentation separately?

Cheers
Stan

williamstein · 2010-04-03T09:10:26Z

comment:14

All three patches still apply fine to sage-4.3.5.

burcin · 2010-04-12T23:18:26Z

comment:19

Thanks for pushing this ticket closer to the finish Dan.

Though I don't think fixing the whitespace problems of sage/symbolic/expression.pyx in attachment: trac_3852_fix_whitespace.patch is a good idea. There are many positively reviewed symbolics tickets waiting to be merged on trac and many of them touch the same file. Pushing the patch on my queue, I get

applying trac_3852_fix_whitespace.patch
patching file sage/symbolic/expression.pyx
Hunk #7 succeeded at 317 with fuzz 1 (offset -2 lines).
Hunk #208 FAILED at 6448
Hunk #215 FAILED at 6699
Hunk #216 FAILED at 6713
3 out of 236 hunks FAILED -- saving rejects to file sage/symbolic/expression.pyx.rej

I don't think rebasing the patch on the other changes is worth it. We should either have an understanding that this ticket is merged after all the other actual bug fixes, and the failures in applying this patch are ignored, or change this patch to fix only the relevant sections of the file. (Actually, I wouldn't mind if this was a general policy on whitespace fixes.)

Otherwise, I'm willing to give Dan's changes and this ticket a positive review.

williamstein · 2010-04-13T03:22:56Z

comment:22

Replying to @dandrake:

attachment: trac_3852_referee.patch adds the units module to the reference manual and makes some minor changes to docstrings. It doesn't change any code.

I'm marking this as "needs review"; if someone could look over my referee patch, we can give this entire ticket a positive review.

I definitely give that referee patch a positive review. Very nice! It's going to be great getting this code into Sage.

burcin · 2010-04-13T10:51:29Z

comment:23

The new white space fix patch (attachment:trac_3852_fix_whitespace.patch) applies cleanly, and all tests pass. I give a positive review to the referee patch (attachment:trac_3852_referee.patch).

Thanks a lot for fixing this Dan!

Patches to be applied are:

jhpalmieri · 2010-04-15T06:02:12Z

comment:24

Merged in 4.4.alpha0:

(I think these patches cause some warning when building the reference manual. Those can be fixed on a follow-up ticket.)

jhpalmieri · 2010-04-15T06:02:12Z

Merged: sage-4.4.alpha0

jasongrout added this to the sage-4.4 milestone Aug 14, 2008

jasongrout added c: numerical labels Aug 14, 2008

jasongrout assigned sagetrac-jkantor Aug 14, 2008

jasongrout unassigned sagetrac-jkantor Aug 14, 2008

jasongrout added c: basic arithmetic and removed c: numerical labels Aug 14, 2008

This comment has been minimized.

Sign in to view

williamstein added s: needs review and removed s: needs work labels Jan 26, 2010

burcin added s: positive review and removed s: needs review labels Apr 13, 2010

jhpalmieri removed the s: positive review label Apr 15, 2010

jhpalmieri closed this as completed Apr 15, 2010

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

create or adapt or include a units package #3852

create or adapt or include a units package #3852

jasongrout commented Aug 14, 2008

jasongrout commented Aug 14, 2008

jasongrout commented Aug 20, 2008

williamstein commented Oct 4, 2009

sagetrac-edrex mannequin commented Nov 4, 2009

sagetrac-edrex mannequin commented Nov 4, 2009

jasongrout commented Nov 4, 2009

sagetrac-edrex mannequin commented Nov 5, 2009

This comment has been minimized.

williamstein commented Jan 21, 2010

sagetrac-schymans mannequin commented Jan 26, 2010

williamstein commented Apr 3, 2010

burcin commented Apr 12, 2010

williamstein commented Apr 13, 2010

burcin commented Apr 13, 2010

jhpalmieri commented Apr 15, 2010

jhpalmieri commented Apr 15, 2010

create or adapt or include a units package #3852

create or adapt or include a units package #3852

Comments

jasongrout commented Aug 14, 2008

jasongrout commented Aug 14, 2008

jasongrout commented Aug 20, 2008

williamstein commented Oct 4, 2009

sagetrac-edrex mannequin commented Nov 4, 2009

sagetrac-edrex mannequin commented Nov 4, 2009

jasongrout commented Nov 4, 2009

sagetrac-edrex mannequin commented Nov 5, 2009

This comment has been minimized.

williamstein commented Jan 21, 2010

sagetrac-schymans mannequin commented Jan 26, 2010

williamstein commented Apr 3, 2010

burcin commented Apr 12, 2010

williamstein commented Apr 13, 2010

burcin commented Apr 13, 2010

jhpalmieri commented Apr 15, 2010

jhpalmieri commented Apr 15, 2010