Diff for "days3/final" - Sagemath Wiki

Differences between revisions 4 and 10 (spanning 6 versions)

1. p-adics

Lesson: SVN + HG only mix together well if you're only creating new files

Status:

Code in place, still unexposed in SAGE, working on more doctests.
Lazy p-adics currently under development
Extension fields are next

Goals: Make (i)-(iii) production ready

Problems: (a) Extensions may be extremely slow and may require FiniteField, PolyQuotientRing optimization to fix.

People: David Roe (DR), David Harvey (DH), Craig Citro (CC), Genya Zaytman (YZ)

Timeline:

Done by AWS: DR/DH/CC/YZ
Done by AWS: DR
Unramified by AWS: CC/DR
All of us will be at AWS

2. Symbolic manipulation

Problems:

Hard to choose when to simplify
Still dependant on Maxima for some things -- will this limit us in the future? (consensus here seems to be probably not)

Status:

Plotting still untouched, esp wrt speed (we think)
We have equality working! Also, formal inequalities ready to go (in principle)

Goals:

More doctests, production ready code
Explore Mathematica, Maple and Maxima, add functionality. Go through a calculus textbook and make sure we can do all the problems.
More user feedback on interface and features
Rouche's Algorithm
Turn function objects/symbolic expressions into C/SageX code.
Make (v) part of a SAGE-wide system

People: Bobby M, Nathan, Soroosh Yazdam, Robert Bradshaw

Timeline:

Before 3/1, Initial Public Release
Get a fair amount of this done by 3/8
March/April (Nathan and Mathematica worksheets)
Soroosh wants to do this when the code base is ready, research other programs by 4/1
April/May
Later

3. Low level optimization of SAGE infrastructure

Goals:

Optimize the code that SageX generates: make Pythonesque syntax fast in lieu of ugly, nasty, Python/C API calls (to avoid coredumps, it's hard to read, easy to write bugs and you feel guilty when you write it)
Speed up SAGE integer construction by changing to handwritten C in SageX
Integer pool?

Problem: In (ii) Tying handwritten C into SageX in a few specific places when SageX is nice but slow. We don't necessarily want to make this easy, and don't know exactly how we should change SageX in order to do this.

Status:

Fairly far along in indexing into lists, but goal (i) is more long term. People can ask why is something slow, look at the generated C code and see if there's a better way to do it. Make a wiki page.
Have code written, trying to get it in place. Understand what's going on. See Problem.
Still thinking about it.

People: Robert Bradshaw, Martin, David Harvey (sick of this project now)

Timeline:

List indexing: 48 hours (Robert). Make a wiki page: more as it comes along.
Figure out problem and implement solution by 3/1
After 2.
Look at schedule for Google summer of code. Python Software Foundation wiki -- describe this project, by 4/1.

4. 3-D plot embedded Java applet

Goal: Make it work and easy to install, even if your name isn't Robert.

Problem: Notebook server chokes on huge files. Maybe rewrite notebook backend in twisted.

People: Tom Boothby

Status: Has twisted server working, but it's not tied in to the notebook client.

5. VMWare

Goal: Sage in a VMWare Player. Compressed VMWare Image should fit on a CD.

People: Marcus, William

Timeline: By AWS

6. Linear Algebra

	Sparse	Dense
GF(2)	(A)	(B)
Numerical	(C)	(D)
Exact	(E)	(F)

A) M4R (Method of Four Russians)

Remark: Do as dense up to certain size, ie A contained in B up to some size

Goal: Get M4R into SAGE

Goal: Get M4R into SAGE.

Example current timings:

	SAGE	Linbox	Bard Gaussian Elim	Bard M4R	Magma
6000x6000		15s	4s	1.7s	2.23s
3000x3000	20s	5s	0.420s	0.2s	0.32s

Status:

Goal: (i) Get M4R into SAGE

(ii) Figure out where to draw the line

People: Martin

Timeline: Once B is done

Goal: Get M4R into SAGE.

Example current timings:

	SAGE	Linbox	Bard Gaussian Elim	Bard M4R	Magma
6000x6000		15s	4s	1.7s	2.23s
3000x3000	20s	5s	0.420s	0.2s	0.32s

People: Martin

Status: Now understand M4R

Timeline: Soon. 3/15

C) Waiting on Suvrit Sra, otherwise clueless. Could do using Suvrit's (or other) libraries if this stalls.

Goal: Implement something to compare to Suvrit

People: Josh

Status: Have a little bit, but it's not very good.

Timeline: Incremental

D) Matrix-vector now BLAS (like everyone)

Using numpy for everything else seems to be the best plan

Goal: Convert

People: Josh

Timeline: 2 weeks

E) (Sparse integer vectors) and (Sparse rational vectors): working, not optimized. Both in progress

(Sparse integer matrices) and (Sparse rational matrices): close

Status: Some benchmarks

Goal:

Rational sparse echelon via multi-modular working
Optimize

Timeline:

72 hours
By 3/10

Remark: Still behind Magma because of mod p, but linbox may help.

F) Status: Lots of benchmarks. Much faster than two weeks ago. Linbox integration coming along nicely.

Remark: Magma may not be producing provably correct answers

Goal:

Better prime selection in multi-modular: maybe 4x speedup
More specialized matrix classes
Optimize matrix-vector and vector-matrix for specialized classes
A lot of optimization (motivated by modular symbols)
Raw methods, much of this will have to be in specialized matrix classes. Maybe not do this (William is not currently convinced)? Look into this.
Examples

Timeline:

By 3/10
Long term

7. Examples

They autotest, done by today -- Rishi

Plan: If you want to put tests in a docstring that the user doesn't see

-  ⇤ ← Revision 4 as of 2007-02-21 20:23:20 → 
  Size: 4058
  Editor: anonymous
  Comment:
+   ← Revision 10 as of 2008-11-14 13:41:50 → ⇥
  Size: 6007
  Editor: anonymous
  Comment: converted to 1.6 markup
-Deletions are marked like this.
+Additions are marked like this.
 Line 4:
-Status: (i) Code in place, still unexposed in SAGE, working on more doctests.
        (ii) Lazy p-adics currently under development
        (iii) Extension fields are next
+Status: 
  1. Code in place, still unexposed in SAGE, working on more doctests.
  2. Lazy p-adics currently under development 
  3. Extension fields are next
-Line 8:
+Line 11:
-Line 9:
+Line 13:
-Line 10:
+Line 15:
-Timeline: (i) Done by AWS: DR/DH/CC/YZ
          (ii) Done by AWS: DR
          (iii) Unramified by AWS: CC/DR
          (iv) All of us will be at AWS
+Timeline: 
   1. Done by AWS: DR/DH/CC/YZ 
   2. Done by AWS: DR
   3. Unramified by AWS: CC/DR
   4. All of us will be at AWS
-Line 16:
+Line 24:
-Problems: (i) Hard to choose when to simplify
          (ii) Still dependant on Maxima for some things -- will this limit us in the future? (consensus here seems to be probably not)
Status: (i) Plotting still untouched, esp wrt speed (we think)
        (ii) We have equality working! Also, formal inequalities ready to go (in principle)
Goals: (i) More doctests, production ready code
       (ii) Explore Mathematica, Maple and Maxima, add functionality.  Go through a calculus textbook and make sure we can do all the problems.
       (iii) More user feedback on interface and features
       (iv) Rouche's Algorithm
       (v) Turn function objects/symbolic expressions into C/SageX code.
       (vi) Make (v) part of a SAGE-wide system
+Problems: 
     1. Hard to choose when to simplify
     2. Still dependant on Maxima for some things -- will this limit us in the future? (consensus here seems to be probably not)

Status: 
     1. Plotting still untouched, esp wrt speed (we think)
     2. We have equality working! Also, formal inequalities ready to go (in principle)

Goals: 
     1. More doctests, production ready code
     2. Explore Mathematica, Maple and Maxima, add functionality.  Go through a calculus textbook and make sure we can do all the problems.
     3. More user feedback on interface and features
     4. Rouche's Algorithm
     5. Turn function objects/symbolic expressions into C/SageX code.
     6. Make (v) part of a SAGE-wide system
-Line 27:
+Line 42:
-Timeline: (i) Before 3/1, Initial Public Release
          (ii) Get a fair amount of this done by 3/8
          (iii) March/April (Nathan and Mathematica worksheets)
          (iv) Soroosh wants to do this when the code base is ready, research other programs by 4/1
          (v) April/May
          (vi) Later
+Timeline: 
     1. Before 3/1, Initial Public Release
     2. Get a fair amount of this done by 3/8
     3. March/April (Nathan and Mathematica worksheets)
     4. Soroosh wants to do this when the code base is ready, research other programs by 4/1
     5. April/May
     6. Later
-Line 35:
+Line 52:
-Goals: (i) Optimize the code that SageX generates: make Pythonesque syntax fast in lieu of ugly, nasty, Python/C API calls (to avoid coredumps, it's hard to read, easy to write bugs and you feel guilty when you write it)
       (ii) Speed up SAGE integer construction by changing to handwritten C in SageX
       (iii) Integer pool?
+Goals: 
     1. Optimize the code that SageX generates: make Pythonesque syntax fast in lieu of ugly, nasty, Python/C API calls (to avoid coredumps, it's hard to read, easy to write bugs and you feel guilty when you write it)
     2. Speed up SAGE integer construction by changing to handwritten C in SageX
     3. Integer pool?
-Line 39:
+Line 59:
-Status: (i) Fairly far along in indexing into lists, but goal (i) is more long term.  People can ask why is something slow, look at the generated C code and see if there's a better way to do it.  Make a wiki page.
        (ii) Have code written, trying to get it in place.  Understand what's going on.  See Problem.
        (iii) Still thinking about it.
+Status: 
      1. Fairly far along in indexing into lists, but goal (i) is more long term.  People can ask why is something slow, look at the generated C code and see if there's a better way to do it.  Make a wiki page.
      2. Have code written, trying to get it in place.  Understand what's going on.  See Problem.
      3. Still thinking about it.
-Line 43:
+Line 66:
-Timeline: (i) List indexing: 48 hours (Robert).  Make a wiki page: more as it comes along.
          (ii) Figure out problem and implement solution by 3/1
          (iii) After (ii)
          (iv) Look at schedule for Google summer of code.  Python Software Foundation wiki -- describe this project, by 4/1.
+Timeline: 
      1. List indexing: 48 hours (Robert).  Make a wiki page: more as it comes along.
      2. Figure out problem and implement solution by 3/1
      3. After 2.
      4. Look at schedule for Google summer of code.  Python Software Foundation wiki -- describe this project, by 4/1.
-Line 49:
+Line 75:
-Line 50:
+Line 77:
-Line 51:
+Line 79:
-Line 52:
+Line 81:
-Line 54:
+Line 84:
-Line 55:
+Line 87:
-Line 56:
+Line 89:
-Line 57:
+Line 91:
-Line 66:
+Line 101:
-Line 67:
+Line 103:
-Line 68:
+Line 105:
-Line 69:
+Line 107:
-Line 70:
+Line 109:
-Line 76:
+Line 116:
+Goal: (i) Get M4R into SAGE

      (ii) Figure out where to draw the line

People: Martin

Timeline: Once B is done


B) Goal: Get M4R into SAGE. 

Example current timings:

|| ||SAGE||Linbox||Bard Gaussian Elim||Bard M4R||Magma||
||6000x6000|| ||15s||4s||1.7s||2.23s||
||3000x3000||20s||5s||0.420s||0.2s||0.32s||

People: Martin

Status: Now understand M4R

Timeline: Soon.  3/15


C) Waiting on Suvrit Sra, otherwise clueless.  Could do using Suvrit's (or other) libraries if this stalls.

Goal: Implement something to compare to Suvrit

People: Josh

Status: Have a little bit, but it's not very good.

Timeline: Incremental

D) Matrix-vector now BLAS (like everyone)

Using numpy for everything else seems to be the best plan

Goal: Convert

People: Josh

Timeline: 2 weeks


E) (Sparse integer vectors) and (Sparse rational vectors): working, not optimized.  Both in progress

   (Sparse integer matrices) and (Sparse rational matrices): close

Status: Some benchmarks

Goal: 
  1. Rational sparse echelon via multi-modular working
  2. Optimize

Timeline: 
  1. 72 hours
  2. By 3/10

Remark: Still behind Magma because of mod p, but linbox may help.



F) Status: Lots of benchmarks.  Much faster than two weeks ago.  Linbox integration coming along nicely.

Remark: Magma may not be producing provably correct answers

Goal: 
  1. Better prime selection in multi-modular: maybe 4x speedup
  2. More specialized matrix classes
  3.  Optimize matrix-vector and vector-matrix for specialized classes
  4. A lot of optimization (motivated by modular symbols)
  5. Raw methods, much of this will have to be in specialized matrix classes.  Maybe not do this (William is not currently convinced)? Look into this.
  6. Examples

Timeline: 
  1. By 3/10
  2. Long term


7. Examples

They autotest, done by today -- Rishi



Plan: If you want to put tests in a docstring that the user doesn't see