Are there any good, up to date books that people recommend for numerical work with Python? I see the book Python Scripting for Computational Science Hans Petter Langtangen Does anyone have opinions on this? Is it current? Are there better books? All pointers appreciated.
On Tue, 2007-02-27 at 21:10 -0600, Robert Love wrote:
Are there any good, up to date books that people recommend for numerical work with Python?
Travis Oliphant's "Guide to NumPy" (eBook) tells you more than you'll probably ever need to know about NumPy specifically, but it's less a recipe book and more a reference manual. I don't know of any, but I'd sure like to see them, since it's a bit of a barrier to entry for new users who see similar books for Matlab, etc. David
On 2/27/07, Robert Love <rblove@airmail.net> wrote:
Are there any good, up to date books that people recommend for numerical work with Python?
I see the book
Python Scripting for Computational Science Hans Petter Langtangen
Does anyone have opinions on this? Is it current? Are there better books?
I specifically do not recommend this book -- I own it but in my opinion it is outdated and is more a collection of the author's personal idioms than the current common practice in the scientific python community. For numerical work in python most people use * numpy - for array math. The best documentation is Travis' online book http://www.tramy.us. For free, the numarray documentation is excellent and the API is very similar for common use cases - http://www.stsci.edu/resources/software_hardware/numarray/doc * scipy - there is no comprehensive printed documentation that I know of. The online help, site docs, and wiki are your best bet. - http://www.scipy.org/Documentation * ipython - the enhanced python shell which is widely used in the scientific computing community for interactive work. Has special modes to support numpy, scipy, matplotlib and distributed computing. - http://ipython.scipy.org/moin/Documentation * matplotlib - 2D graphics, charts and the like. ipython has a 'ipython -pylab' mode which loads matplotlib and numpy for a matlab like environment . http://matplotlib.sourceforge.net/tutorial.html and http://matplotlib.sourceforge.net/users_guide_0.87.7.pdf * Enthought Tool Suite - provides a comprehensive package for scientific computing including the above modules and many others for application development and more. Provides 3D graphics through the Mayavi2/VTK packages and 2D plotting through Chaco, and lots of tools to facilitate wx based GUI development - http://code.enthought.com/ There is a lot more, particularly for domain specific stiff, but these links are good starting points. Unfortunately, there is no one-stop-shop for a guide to scientific computing in python - Travis' documentation is the closest thing we have but it pretty much just covers numpy which is *the* core package. Fernando Perez and I have a very brief and limited started guide covering multiple packages (ipython, numpy, matplotlib, scipy, VTK) but I don't have the PDF handy (Fernando, do you have the roadshow doc handy?). Eric Jones and Travis (authors of scipy) have some talk notes at http://www.nanohub.org/resources/?id=99 but these are a bit out of date. JDH
On 2/27/07, John Hunter <jdh2358@gmail.com> wrote:
On 2/27/07, Robert Love <rblove@airmail.net> wrote:
Are there any good, up to date books that people recommend for numerical work with Python?
I see the book
Python Scripting for Computational Science Hans Petter Langtangen
Does anyone have opinions on this? Is it current? Are there better books?
I specifically do not recommend this book -- I own it but in my opinion it is outdated and is more a collection of the author's personal idioms than the current common practice in the scientific python community. For numerical work in python most people use
I happen to share John's opinion, and I also have a copy of this book. While it's technically correct, well written and fairly comprehensive (probably /too/ much, since it's a bit all over the map), I strongly dislike his approach. Much of the book uses his custom, home-made collection of scripts and tools, which you can only download if you go to a site and type a word from a certain page in the book (a simple 'protection' system). Now you have an unmaintained, unreleased (publicly), set of tools to learn from that don't have any licensing explicitly specified. Oh, and a good chunk of the tools in his distribution (since I have the book, I have the code) use Perl. Go figure (there's also a tcl directory thrown in for good measure). One of Python's main strengths for scientific work is precisely the openness and interoperability of the various tools, and we all do our part to help that be the case. The fact that this book follows an approach more or less orthogonal to those ideas makes me very much uninterested in using it.
There is a lot more, particularly for domain specific stiff, but these links are good starting points. Unfortunately, there is no one-stop-shop for a guide to scientific computing in python - Travis' documentation is the closest thing we have but it pretty much just covers numpy which is *the* core package. Fernando Perez and I have a very brief and limited started guide covering multiple packages (ipython, numpy, matplotlib, scipy, VTK) but I don't have the PDF handy (Fernando, do you have the roadshow doc handy?).
Well, you asked for it :) http://amath.colorado.edu/faculty/fperez/tmp/py4science.pdf It's worth stressing, in the strongest possible terms, that this should NOT, in any way, shape or form, be considered anything beyond a pre-pre-alpha, pre-draft of a project for a possible book :) Besides, it's already outdated in several important places (numpy, mayavi, no TVTK,...). After all I said about the Langtangen book, at least it's a real one. Our pdf draft is most certainly not. So if you need a book, with all of its limitations, Langtangen's is currently the only game in town that covers the whole spectrum of python for scientific computing. If John and I ever end up stranded on a desert island for 3 months with great internet access and poor diving gear, we might actually finish ours, but don't hold your breath. Honestly, I think that today your best bet is: 1. Buy Travis' book. It's fantastic, has everything you need to know about numpy, and you'll be supporting numpy itself. 2. Print Perry Greenfield's tutorial (http://new.scipy.org/wikis/topical_software/Tutorial). I think he's updating it for numpy now. 3. Have a look at some of the other info in http://new.scipy.org/Documentation, in particular D. Kuhlman's course is very nice. Regards, f
On Tue, Feb 27, 2007 at 11:19:51PM -0700, Fernando Perez wrote:
Nice ! Really nice ! I am very happy that someone is writing something about all this in a consistent way. I have two questions: what will be the distribution mode (printed matter, internet and free, both) ? This is important because an easily and freely available book really helps spreading the word. Second question: what will be the licence ? The reason I ask this is that I could consider trying to contribute if I can reuse the part I contribute. As you are probably aware (I think you follow the enthought-dev ML) Prabhu and I are currently trying to find time to code a nice "mlab" interface to mayavi2. Once it has made some progress (expect two to three months) it will be much nicer than the current way of using mayavi from python. Of course there is the problem that mayavi2 is not properly distributed currently. That might change as enthought is currently pushing for a modular eggs-based version of their tool suite. Anyway if we can get something out of mayavi's mlab, I could write document how to use it and contribute a modified version of it to your book. Similarly you could use a modified version of my traitsUI tutorial http://www.gael-varoquaux.info/computers/traits_tutorial/index.html if you think this is not out of scope. Thumbs up for such work ! Gaël
Damn it, I intended to send this e-mail off list ! On Wed, Feb 28, 2007 at 08:29:32AM +0100, Gael Varoquaux wrote:
On Tue, Feb 27, 2007 at 11:19:51PM -0700, Fernando Perez wrote:
Nice ! Really nice ! I am very happy that someone is writing something about all this in a consistent way.
I have two questions: what will be the distribution mode (printed matter, internet and free, both) ? This is important because an easily and freely available book really helps spreading the word. Second question: what will be the licence ? The reason I ask this is that I could consider trying to contribute if I can reuse the part I contribute.
As you are probably aware (I think you follow the enthought-dev ML) Prabhu and I are currently trying to find time to code a nice "mlab" interface to mayavi2. Once it has made some progress (expect two to three months) it will be much nicer than the current way of using mayavi from python. Of course there is the problem that mayavi2 is not properly distributed currently. That might change as enthought is currently pushing for a modular eggs-based version of their tool suite.
Anyway if we can get something out of mayavi's mlab, I could write document how to use it and contribute a modified version of it to your book.
Similarly you could use a modified version of my traitsUI tutorial http://www.gael-varoquaux.info/computers/traits_tutorial/index.html if you think this is not out of scope.
Thumbs up for such work !
Gaël
On 2/28/07, Gael Varoquaux <gael.varoquaux@normalesup.org> wrote:
On Tue, Feb 27, 2007 at 11:19:51PM -0700, Fernando Perez wrote:
I'll reply on-list since you sent it on-list :) But feel free to ask off-list anything in particular.
Nice ! Really nice ! I am very happy that someone is writing something about all this in a consistent way.
Thanks, but it's unfortunately a bit of a stillborn effort at this point.
I have two questions: what will be the distribution mode (printed matter, internet and free, both) ? This is important because an easily and freely available book really helps spreading the word. Second question: what will be the licence ? The reason I ask this is that I could consider trying to contribute if I can reuse the part I contribute.
Well, nothing had been set in stone yet, but I'd favor a combination of a printed book by a publisher with some track record in science, hopefully with an agreement that would allow free online redistribution. Such things have been done recently, so perhaps convincing a publisher of this might not be altogether impossible.
As you are probably aware (I think you follow the enthought-dev ML) Prabhu and I are currently trying to find time to code a nice "mlab" interface to mayavi2. Once it has made some progress (expect two to three months) it will be much nicer than the current way of using mayavi from python. Of course there is the problem that mayavi2 is not properly distributed currently. That might change as enthought is currently pushing for a modular eggs-based version of their tool suite.
I know, I've been quietly following that discussion with the utmost interest. I'm hoping it will be ready by Saturday, b/c I'm having a mini-sprint at my house with some students to port some old CFD code to the new mayavi. So you have 3 days to finish it :)
Anyway if we can get something out of mayavi's mlab, I could write document how to use it and contribute a modified version of it to your book.
Similarly you could use a modified version of my traitsUI tutorial http://www.gael-varoquaux.info/computers/traits_tutorial/index.html if you think this is not out of scope.
Thumbs up for such work !
Thanks! But as I said, while I'm quite open on the possibilities for licensing and distribution, the grim reality is that right now, I simply don't have /any/ time to put into this. All of my 'free' (funny, I know) time right now has to go into pushing the new code effort for distributed and parallel computing in ipython. We've made good progress recently on that front and some of the things that were holding back integration of the trunk into the dev branch, so that we might finally have a single codebase. So while it would be very nice for this to be finished, realistically I don't see that coming from me in the near future. Regards, f
On Wed, Feb 28, 2007 at 01:22:34AM -0700, Fernando Perez wrote:
As you are probably aware (I think you follow the enthought-dev ML) Prabhu and I are currently trying to find time to code a nice "mlab" interface to mayavi2. Once it has made some progress (expect two to three months) it will be much nicer than the current way of using mayavi from python. Of course there is the problem that mayavi2 is not properly distributed currently. That might change as enthought is currently pushing for a modular eggs-based version of their tool suite.
I know, I've been quietly following that discussion with the utmost interest. I'm hoping it will be ready by Saturday, b/c I'm having a mini-sprint at my house with some students to port some old CFD code to the new mayavi. So you have 3 days to finish it :)
Well well. Currently the API is not stable at all, and I do not want to freeze it (cf my last about this on the enthougt-dev ML). I can let this project steel even more sleep from me than my work already does, and move a bit forward to try and establish something that is closer to the API that I am aiming to, but even if I a manage to get something out (and when I am to tired I am not terribly productive), I need Prabhu to review it. All I can say is that mayavi.tools.mlab is not currently in its final form and I do not suggest using it, unless you are planning to modify the code you are writing. From the user point of view following the API change should not be a huge amount of work. The changes I would like do to the API are exposed in the mails: https://mail.enthought.com/pipermail/enthought-dev/2007-February/004425.html [[Enthought-dev] Mlab API, and usability] and https://mail.enthought.com/pipermail/enthought-dev/2007-February/004442.html [Enthought-dev] Mlab: contour3d and quiver3d. I am interested on feedback on these propositions, by the way. Gaël
Fernando Perez wrote:
On 2/27/07, John Hunter <jdh2358@gmail.com> wrote:
Are there any good, up to date books that people recommend for numerical work with Python?
I see the book
Python Scripting for Computational Science Hans Petter Langtangen
Does anyone have opinions on this? Is it current? Are there better books? I specifically do not recommend this book -- I own it but in my opinion it is outdated and is more a collection of the author's
On 2/27/07, Robert Love <rblove@airmail.net> wrote: personal idioms than the current common practice in the scientific python community. For numerical work in python most people use
I happen to share John's opinion, and I also have a copy of this book. While it's technically correct, well written and fairly comprehensive (probably /too/ much, since it's a bit all over the map), I strongly dislike his approach. Much of the book uses his custom, home-made collection of scripts and tools, which you can only download if you go to a site and type a word from a certain page in the book (a simple 'protection' system).
Concur with John and Fernando. I found a copy in a local bookstore and bought it because it _is_ the only book covering the subject and I wanted to show that there is demand for such material. # Steve
On Wednesday 28 February 2007 06:29:38 am Steven H. Rogers wrote:
Fernando Perez wrote:
On 2/27/07, John Hunter <jdh2358@gmail.com> wrote:
On 2/27/07, Robert Love <rblove@airmail.net> wrote:
Are there any good, up to date books that people recommend for numerical work with Python?
I see the book
Python Scripting for Computational Science Hans Petter Langtangen
Does anyone have opinions on this? Is it current? Are there better books?
I specifically do not recommend this book -- I own it but in my opinion it is outdated and is more a collection of the author's personal idioms than the current common practice in the scientific python community. For numerical work in python most people use
I happen to share John's opinion, and I also have a copy of this book. While it's technically correct, well written and fairly comprehensive (probably /too/ much, since it's a bit all over the map), I strongly dislike his approach. Much of the book uses his custom, home-made collection of scripts and tools, which you can only download if you go to a site and type a word from a certain page in the book (a simple 'protection' system).
Concur with John and Fernando. I found a copy in a local bookstore and bought it because it _is_ the only book covering the subject and I wanted to show that there is demand for such material.
I did the same thing, and came to the same conclusion about the book. I haven't seen anyone mention "Numerical Methods in Engineering with Python" by Jaan Kiusalaas. It was published in 2005, and uses numarray, so it is somewhat dated considering all the impressive developments with NumPy and SciPy. I mention it for the sake of completeness. Darren
I'm setting up a new site about chemometrics hopefully (cross fingered), it should be online on Thursday. With other co-author we are also thinking about writing a "chapter of a book"/article about how to use python/scipy to perform common chemometric tasks. Why dont' we all merge our knowledge in different field and try to have something more "comprehensive" ? prbably togheter we can succeed in seeing all in completion without all rewriting twice same things We will also be glade to give to scipy the functions to seem them included if anyone interested just drop me a line Giorgio
Darren Dale wrote:
I haven't seen anyone mention "Numerical Methods in Engineering with Python" by Jaan Kiusalaas. It was published in 2005, and uses numarray, so it is somewhat dated considering all the impressive developments with NumPy and SciPy. I mention it for the sake of completeness.
This book has a first chapter on the basics of Python. The remainder is a light weight introduction to numerical methods and does not teach anything about Python. --
On Thu, 2007-03-01 at 21:06 +0000, Donald Fredkin wrote:
I haven't seen anyone mention "Numerical Methods in Engineering with Python" by Jaan Kiusalaas. It was published in 2005, and uses numarray, so it is somewhat dated considering all the impressive developments with NumPy and SciPy. I mention it for the sake of completeness.
This book has a first chapter on the basics of Python. The remainder is a light weight introduction to numerical methods and does not teach anything about Python.
A lab mate of mine showed me this today. It does seem useful in that it contains lots of little recipes. Unfortunately, I think that the recipe I was interested in, namely the conjugate gradient solver linear systems, was incorrect or at least incompatible with NumPy (the residuals were *diverging*, somehow) and I ended up rewriting it from scratch. David
On Feb 28, 2007, at 1:19 AM, Fernando Perez wrote:
2. Print Perry Greenfield's tutorial (http://new.scipy.org/wikis/topical_software/Tutorial). I think he's updating it for numpy now.
Yes, that's right. Hopefully it won't take too long to do (a week or two depending on my other work). Perry
While I agree with some of the criticisms of the Langtangen book I do think it covers some topics that are really hard to find elsewhere. He does go into depth and cover a lot of writing C/C++ extensions. Overall the book has a lot of information. Thumbs down on parts that rely on his own code, especially the Perl stuff, but Thumbs Up on parts that show you details you have a hard time finding elsewhere. The biggest problem with the book is one that ALL printed documents face with Python (and probably other open source projects): staying current. Things change fast and books go out of date fast. Web sources and discussion list like this one are essential. The way I see it as a user of scientific python packages the biggest barriers to new users and even veterans are: * Getting documention, examples and tutorials for various package. Can vary greatly depending on each package. * Figuring out how to install packages (an on-going battle for all) * Figuring out package dependencies (e.g. do I need wxPython with matplotlib?) * Figuring out which versions of each package are compatible with other packages' versions. I bet I'm not the only one who gets nervous when he/she has to update a package. --- Fernando Perez <fperez.net@gmail.com> wrote:
On 2/27/07, John Hunter <jdh2358@gmail.com> wrote:
On 2/27/07, Robert Love <rblove@airmail.net> wrote:
Are there any good, up to date books that people recommend for numerical work with Python?
I see the book
Python Scripting for Computational Science Hans Petter Langtangen
Does anyone have opinions on this? Is it current? Are there better
-- Lou Pecora, my views are my own. --------------- Three laws of thermodynamics: First law: "You can't win." Second law: "You can't break even." Third law: "You can't quit." -- Allen Ginsberg, beat poet ____________________________________________________________________________________ Expecting? Get great news right away with email Auto-Check. Try the Yahoo! Mail Beta. http://advision.webevents.yahoo.com/mailbeta/newmail_tools.html
I would help write part of a book. I'm sure others would, too. But none of us have the time to write a whole book ourselves. Who could coordinate such an effort? What format would be used (LaTeX, etc.)? How would consistency be maintained across sections? I think these are not insurmountable obstacles, and a free, printable PDF howto manual (as opposed to a reference book like the NumPy book) in the spirit of the Lerning ... O'Reilly books is very, very important. The tutorials on the wiki are a good place to start, but in the end, people want a book. I know I prefer books to online documentation when I am really diving into something. -Rob ---- Rob Hetland, Associate Professor Dept. of Oceanography, Texas A&M University http://pong.tamu.edu/~rob phone: 979-458-0096, fax: 979-845-6331
On 2/28/07, Rob Hetland <hetland@tamu.edu> wrote:
I would help write part of a book. I'm sure others would, too. But none of us have the time to write a whole book ourselves.
Who could coordinate such an effort?
What format would be used (LaTeX, etc.)?
How would consistency be maintained across sections?
I think these are not insurmountable obstacles, and a free, printable PDF howto manual (as opposed to a reference book like the NumPy book)
When Fernando and I first started the project, we wanted to keep this with as few authors as possible simply because most team written books aren't that good, and are not well integrated. Maintenance also becomes difficult since open source is a rapidly moving target (as Fernando noted several of our chapters are already out of date) and the more authors you have the more difficult it becomes to herd the cats. So we decided to try and tackle it alone. Unfortunately, both of us are too involved with our other commitments to put the work in that is necessary, and speaking for myself, I would be happy to revisit the idea of a collaborative project. We would need a couple of people to step up as editors who commit a fair amount of time to insuring style, cross referencing, proper topics, etc... We could then solicit chapters on all the relevant major packages, either from the authors of the packages or from heavy users. We would need a few introductory chapters which emphasize integrated usage, followed by package specific chapters emphasizing deeper features. At this point, I can probably only commit to an mpl chapter and maybe some work on some integration. I would be happy to do this under an open document license, while still striving for hardcopy printing. The src for the document Fernando posted lives in matplotlib svn as a collection of lyx chapters (egad the last commit was 20 months ago) http://matplotlib.svn.sourceforge.net/viewvc/matplotlib/trunk/course/ I will defer to Fernando's wishes on this, but I am happy to try and move the ball forward by bringing in any one offering to do some work. Perry's tutorial is very nice and covers a lot of ground, often in more detail than what Fernando and I have done, and might be a better starting point, depending on his interests. Or some variant of his tutorial might serve well for the introductory chapters, followed by the package specific stuff. JDH
On Feb 28, 2007, at 12:10 PM, John Hunter wrote:
Perry's tutorial is very nice and covers a lot of ground, often in more detail than what Fernando and I have done, and might be a better starting point, depending on his interests. Or some variant of his tutorial might serve well for the introductory chapters, followed by the package specific stuff.
I see it as suitable for a certain class of new users (i.e, those that are more task oriented and looking to see what they can accomplish interactively without having to learn a lot of preliminary material. But it certainly isn't suitable for all. I was talking to Travis about the need for some sort of lightweight intro to numpy. His book is a great technical reference, and very suitable for developers. But I worry that it is a bit intimidating for new users. So I raised the possibility of writing some sort of introduction that did the basics but pointed to the book for details for more advanced topics. Travis was ok with that (I didn't want to instigate anything that would cut into book sales). I was thinking of perhaps taking the old Numeric/numarray user guides and stripping them down and modifying that base document after I finish with the interactive tutorial (if I have time of course). With having to switch our staff over to numpy (we've pretty much completed converting most of our distributed software; our next release in early summer will be numpy-based) I think we need something that isn't too long to read. If others want to take this on, I welcome it (and as I said, I don't think Travis objects based on what he told me). Perry
What's really is needed are Examples, examples, and examples. The book is light on that (no offense). Some defintions of functions or methods leaves me scratching my head asking, "How is that done in the code, really?" --- Perry Greenfield <perry@stsci.edu> wrote:
I see it as suitable for a certain class of new users (i.e, those that are more task oriented and looking to see what they can accomplish interactively without having to learn a lot of preliminary material. But it certainly isn't suitable for all.
I was talking to Travis about the need for some sort of lightweight intro to numpy. His book is a great technical reference, and very suitable for developers. But I worry that it is a bit intimidating for new users. So I raised the possibility of writing some sort of introduction that did the basics but pointed to the book for details for more advanced topics. Travis was ok with that (I didn't want to instigate anything that would cut into book sales). I was thinking of perhaps taking the old Numeric/numarray user guides and stripping them down and modifying that base document after I finish with the interactive tutorial (if I have time of course). With having to switch our staff over to numpy (we've pretty much completed converting most of our distributed software; our next release in early summer will be numpy-based) I think we need something that isn't too long to read. If others want to take this on, I welcome it (and as I said, I don't think Travis objects based on what he told me).
Perry
-- Lou Pecora, my views are my own. --------------- Three laws of thermodynamics: First law: "You can't win." Second law: "You can't break even." Third law: "You can't quit." -- Allen Ginsberg, beat poet ____________________________________________________________________________________ Finding fabulous fares is fun. Let Yahoo! FareChase search your favorite travel sites to find flight and hotel bargains. http://farechase.yahoo.com/promo-generic-14795097
On Wed, 28 Feb 2007, Lou Pecora wrote:
What's really is needed are Examples, examples, and examples. The book is light on that (no offense). Some defintions of functions or methods leaves me scratching my head asking, "How is that done in the code, really?"
I'll second that suggestion. The mail list is very helpful, but it would be nice to see the actual use of a function in a small program (or function) in the book. Rich -- Richard B. Shepard, Ph.D. | The Environmental Permitting Applied Ecosystem Services, Inc. | Accelerator(TM) <http://www.appl-ecosys.com> Voice: 503-667-4517 Fax: 503-667-8863
I'll second that suggestion. The mail list is very helpful, but it would be nice to see the actual use of a function in a small program (or function) in the book.
Something like that ? http://www.scipy.org/Numpy_Example_List_With_Doc http://www.scipy.org/Cookbook
On Wed, 28 Feb 2007, Pierre GM wrote:
Something like that ? http://www.scipy.org/Numpy_Example_List_With_Doc http://www.scipy.org/Cookbook
Yes. -- Richard B. Shepard, Ph.D. | The Environmental Permitting Applied Ecosystem Services, Inc. | Accelerator(TM) <http://www.appl-ecosys.com> Voice: 503-667-4517 Fax: 503-667-8863
Well... yes, what took you so long? :-) Thanks. A very good web site. And right there hiding under my nose. --- Pierre GM <pgmdevlist@gmail.com> wrote:
Something like that ? http://www.scipy.org/Numpy_Example_List_With_Doc http://www.scipy.org/Cookbook
-- Lou Pecora, my views are my own. --------------- Three laws of thermodynamics: First law: "You can't win." Second law: "You can't break even." Third law: "You can't quit." -- Allen Ginsberg, beat poet ____________________________________________________________________________________ Need Mail bonding? Go to the Yahoo! Mail Q&A for great tips from Yahoo! Answers users. http://answers.yahoo.com/dir/?link=list&sid=396546091
On Wednesday 28 February 2007 15:03:11 Lou Pecora wrote:
Well... yes, what took you so long? :-)
Thanks. A very good web site. And right there hiding under my nose.
If I remember my Poe correctly, that's quite often the case. But seriously: I find the Numpy examples invaluable. However, nothing prevents us to start reorganizing these examples in a fashion closer to Moler's approach: short, real-life examples, along with the corresponding Numpy code. A kind of Guide to numpy's electronic companion, that shouldn't replace the Cookbook but complete it. We should first to agree on a basic template, and fill it adequately. The wiki organization is ideal for that.
--- Pierre GM <pgmdevlist@gmail.com> wrote:
On Wednesday 28 February 2007 15:03:11 Lou Pecora wrote:
Well... yes, what took you so long? :-)
Thanks. A very good web site. And right there hiding under my nose.
If I remember my Poe correctly, that's quite often the case.
You know Edgar Allen well. I think that was "The Tell-Tale Heart". Very good knowledge of literature.
But seriously: I find the Numpy examples invaluable. However, nothing prevents us to start reorganizing these examples in a fashion closer to Moler's approach: short, real-life examples, along with the corresponding Numpy code. A kind of Guide to numpy's electronic companion, that shouldn't replace the Cookbook but complete it. We should first to agree on a basic template, and fill it adequately. The wiki organization is ideal for that.
You are right. I'm not sure what the format should be. I'm a little out of it about writing up things even though I have contributed to the NumPy/SciPy wiki. -- Lou Pecora, my views are my own. --------------- Three laws of thermodynamics: First law: "You can't win." Second law: "You can't break even." Third law: "You can't quit." -- Allen Ginsberg, beat poet ____________________________________________________________________________________ No need to miss a message. Get email on-the-go with Yahoo! Mail for Mobile. Get started. http://mobile.yahoo.com/mail
On 2/28/07, Lou Pecora <lou_boog2000@yahoo.com> wrote:
--- Pierre GM <pgmdevlist@gmail.com> wrote:
Thanks. A very good web site. And right there hiding under my nose.
If I remember my Poe correctly, that's quite often the case.
You know Edgar Allen well. I think that was "The Tell-Tale Heart". Very good knowledge of literature.
I suspect the reference was to "The Purloined Letter" (or "La carta robada", as I originally read it) instead. Cheers, f
--- Fernando Perez <fperez.net@gmail.com> wrote:
You know Edgar Allen well. I think that was "The Tell-Tale Heart". Very good knowledge of
On 2/28/07, Lou Pecora <lou_boog2000@yahoo.com> wrote: literature.
I suspect the reference was to "The Purloined Letter" (or "La carta robada", as I originally read it) instead.
Sigh. This is why I went into physics and not literature. Thanks. :-) -- Lou Pecora, my views are my own. --------------- Three laws of thermodynamics: First law: "You can't win." Second law: "You can't break even." Third law: "You can't quit." -- Allen Ginsberg, beat poet ____________________________________________________________________________________ Have a burning question? Go to www.Answers.yahoo.com and get answers from real people who know.
Personally, I think the approach Cleve Moler ( of Mathworks) takes in the first chapter of "Numerical Computing in Matlab" is the best I have seen. http://www.mathworks.com/moler/chapters.html What he does is, he takes a central concept ( Fibonacci numbers and Golden Ratio). Explains everything about them using interactive examples in Matlab-scripts. This approach helps a reader to learn when to use a function and how to use it. So if one can explain a lightweight scientific-computing concept using Scipy / Numpy and exhausts most of the functions that would be a great tutorial. Just my 2 cents. --Krish Subramaniam On 2/28/07, Rich Shepard <rshepard@appl-ecosys.com> wrote:
On Wed, 28 Feb 2007, Lou Pecora wrote:
What's really is needed are Examples, examples, and examples. The book is light on that (no offense). Some defintions of functions or methods leaves me scratching my head asking, "How is that done in the code, really?"
I'll second that suggestion. The mail list is very helpful, but it would be nice to see the actual use of a function in a small program (or function) in the book.
Rich
-- Richard B. Shepard, Ph.D. | The Environmental Permitting Applied Ecosystem Services, Inc. | Accelerator(TM) <http://www.appl-ecosys.com> Voice: 503-667-4517 Fax: 503-667-8863 _______________________________________________ SciPy-user mailing list SciPy-user@scipy.org http://projects.scipy.org/mailman/listinfo/scipy-user
On 2007-02-28, Fernando Perez <fperez.net@gmail.com> wrote:
Well, you asked for it :)
Cool. If you're interested in suggestions for additional topics, you might want to add a chapter on scientific python: http://sourcesup.cru.fr/projects/scientific-py/ Though it's probably not "cool" these days compared to some of the other visualization options, I still get a _lot_ of use out of gnuplot-py: http://gnuplot-py.sourceforge.net/ -- Grant Edwards grante Yow! Civilization is at fun! Anyway, it keeps visi.com me busy!!
participants (16)
-
Darren Dale
-
David Warde-Farley
-
Donald Fredkin
-
Fernando Perez
-
Gael Varoquaux
-
Giorgio Luciano
-
Grant Edwards
-
John Hunter
-
Krish Subramaniam
-
Lou Pecora
-
Perry Greenfield
-
Pierre GM
-
Rich Shepard
-
Rob Hetland
-
Robert Love
-
Steven H. Rogers