Re: comparing two lists and returning position
hiro wrote: bare in mind that I have a little over 10 million objects in my list (l2) and l1 contains around 4 thousand objects.. (i have enough ram in my computer so memory is not a problem) Glad to see you solved the problem with the trailing space. Just one minor point, I did say or probably less efficiently ... As far as i know, my suggestion's running time is proportional to len(l1)*len(l2), which gets quite big for your case where l1 and l2 are large lists. If I understand how python dictionaries work, Paul Rubin's suggestion from itertools import izip, count pos = map(dict(izip(l2, count())).__getitem__, l1) or the (I think) approximately equivalent from itertools import izip, count d = dict(izip(l2,count())) pos = [ d[i] for i in l1 ] or the more memory intensive d = dict(zip(l2,range(len(l2 pos = [ d[i] for i in l1 ] should all take take running time proportional to (len(l1)+len(l2))*log(len(l2)) For len(l1)=4,000 and len(l2)=10,000,000 Paul's suggestion is likely to take about 1/100th of the time to run, ie be about 100 times as fast. I was trying to point out a somewhat clearer and simpler (but slower) alternative. Charles -- http://mail.python.org/mailman/listinfo/python-list
Re: comparing two lists and returning position
Paul Rubin wrote: from itertools import izip pos = map(dict(izip(l2, count())).__getitem__, l1) or probably less efficiently ... l1 = [ 'abc', 'ghi', 'mno' ] l2 = [ 'abc', 'def', 'ghi', 'jkl', 'mno', 'pqr'] pos = [ l2.index(i) for i in l1 ] print pos [0, 2, 4] Charles -- http://mail.python.org/mailman/listinfo/python-list
Re: ctypes: error passing a list of str to a fortran dll
luis wrote: I'm using ctypes to call a fortran dll from python. I have no problems passing integer and double arryas, but I have an error with str arrys. For example: [snip] I do not know about Microsoft Fortran compilers (your mention of dll indicates you are probably using MS), nor much about Python, but the C equivalent of a given Fortran call is operating system and compiler dependent. You should consult the Fortran compiler manual for the compiler used to create the DLL. Despite this, many (but not all) C to Fortran interfaces have the following characteristics + C name is Fortran name in lower case + Fortran REAL, DOUBLE PRECISION, INTEGER etc parameters are pointers to the parameter in C, ie float*, etc + Fortran REAL etc arrays are pointers in C (same as C arrays decay to pointers). + Fortran CHARACTER and CHARACTER arrays are passed as TWO parameters, one a pointer to the start of the variable or array, and the other the length as an integer, not a pointer. The length parameters follow all other parameters, in order of the character variables/arrays. Variations I have seen (all on Unix) include using upper case instead of lower, prepending (or postpending) an underscore (or other character(s)) to the subroutine or function name, and using special character descriptors (packing address and length into one word) for character variables. There are almost certainly more variations that I have not seen. For example, given a FORTRAN declaration SUBROUTINE X( CV, RV, CA, N ) CHARACTER*(*) CV REAL RV CHARACTER*(*) CA(*) INTEGER N The C equivalent is likely to be void x( char *pcv, float *prv, char *pca, int *pn, int lv, int la) Where lv will hold the length of cv and la the length of each element of ca (required to be the same for all elements of the array). Fortran uses fixed length character strings, padded with blanks. Given the error message ValueError: Procedure probably called with not enough arguments (4 bytes missing) I suspect that your Fortran compiler is one of the many which do this, and the missing 4 bytes are the integer length of each element of the character array. Also, I noticed you seem to be passing an array of character pointers rather than an array of characters. It is doubtful that Fortran can handle this. You will probably have to pad the strings to a maximal length with spaces, concatanate then into one big string, and pass this by reference together with their padded length. Your Fortran may (but probably won't) have extensions that allow you to pass an array of character pointers. Charles -- http://mail.python.org/mailman/listinfo/python-list
Re: Is PEP-8 a Code or More of a Guideline?
Dave Hansen wrote: The date is about right (actually, a little early: ASR-33, 1965; C, about 1970), but you can't program C on an ASR-33. Keywords are all lower case, and always have been. IF is a syntax error... But the Unix terminal drivers of the day for upper case only terminals translated to lower case and you had to enter \A to get an upper case A, and so on. Still hangs around in the stty options iuclc, -iuclc, olcuc and -olcuc - You can make an x-term window look like an old upper case only terminal. I don't know if any still do it, but old unices used to detect on logon if you typed your user name and password in all caps and then turned on the iuclc and olcuc options, assuming that the reason for the all caps was a upper case only terminal - great fun if you hit caps lock by accident. Charles -- http://mail.python.org/mailman/listinfo/python-list
Re: installing cx_Oracle.
Doug Phillips wrote: It also works the other way around, at least on the non-empty set of systems that contains my workstation. export simply marks the variable name for automatic export to the environment of subsequent commands. The value at that time doesn't matter. What matters is the value that the name has at the time the command is run: [snip] Just tried on a FreeBSD 6.1 development box with stock /bin/sh and it works there too... As far as I know, it has been like this since the introduction of the Bourne shell in (according to a web search) 1979. Charles -- http://mail.python.org/mailman/listinfo/python-list
Re: Regexes: How to handle escaped characters
Torsten Bronger wrote: Hallöchen! [...] Example string: uHollo, escaped positions: [4]. Thus, the second o is escaped and must not be found be the regexp searches. Instead of re.search, I call the function guarded_search(pattern, text, offset) which takes care of escaped caracters. Thus, while Tschö, Torsten. I'm still pretty much a beginner, and I am not sure of the exact requirements, but the following seems to work for at least simple cases when overlapping matches are not considered. def guarded_search( pattern, text, exclude ): return [ m for m in re.finditer(pattern,text) if not [ e for e in exclude if m.start() = e m.end() ] ] txt = axbycz exc = [ 3 ] # y pat = [xyz] mtch = guarded_search(pat,txt,exc) print Guarded search text='%s' excluding %s % ( txt,exc ) for m in mtch: print m.group(), 'at', m.start() txt = Hollo exc = [ 4 ] # Final o pat = o$ mtch = guarded_search(pat,txt,exc) print Guarded search text='%s' excluding %s %s matches % (txt,exc,len(mtch)) for m in mtch: print m.group(), 'at', m.start() Guarded search text='axbycz' excluding [3] 2 matches x at 1 z at 5 Guarded search text='Hollo' excluding [4] 0 matches Simply finds all the (non-overlapping) matches and rejects any that include one of the excluded columns (the y in the first case and the final o in the second). Charles -- http://mail.python.org/mailman/listinfo/python-list
Re: preferred windows text editor?
Ant wrote: What method of executing code snippets in a Python shell do other Vim users use? Other than just copy/paste? Not vim, but good old vi so should work in vim 1. Mark the start of the fragment, for exampls ms (to mark with label s). Labels a through z are available. 2. Move to the end of the fragment. 3. :'s,.w !python to send the fragment to the python interpreter Worked for me when I tried it a few minutes ago. I had never bothered before - just copied/pasted. Obviously, you can also mark the end, move to the start and do something like :.,'ew !python or mark both the start and the end, or use line numbers, or labels plus offsets, or searches, eg :/def/;+5w !python to search forward to the next occurrence of def and send that line plus the next five to the interpreter. Whatever works for you. The abbreviate and map commands can be used to reduce the typing if you are fanatical. Charles -- http://mail.python.org/mailman/listinfo/python-list
Re: interesting exercise
Michael Tobis wrote:
Here is the bloated mess I came up with. I did see that it had to be
recursive, and was proud of myself for getting it pretty much on the
first try, but the thing still reeks of my sorry old fortran-addled
mentality.
Recursion is not necessary, but is much, much clearer.
Here is one non-recursive version from another aging
fortran programmer. I agree it is less clear than most
of the recursive alternatives. No checks for sorted
input etc, these are left as an exercise for the reader.
def permute( s, n ):
def _perm( m, n ):
ilist = [0]*n
while True:
yield ilist
i = n-1
while i = 0 and ilist[i]=m-1: i = i - 1
if i = 0:
ilist = ilist[0:i] + [ilist[i]+1] + [0]*(n-i-1)
else:
return
return [ ''.join([s[i] for i in ilist])
for ilist in _perm(len(s),n) ]
print permute('abc',2) = , permute('abc',2)
print len(permute('13579',3)) = , len(permute('13579',3))
permute('abc',2) = ['aa', 'ab', 'ac', 'ba', 'bb', 'bc',
'ca', 'cb', 'cc']
len(permute('13579',3)) = 125
or even this monstrosity ...
def permute2( s, n ):
return [ ''.join([ s[int(i/len(s)**j)%len(s)]
for j in range(n-1,-1,-1)])
for i in range(len(s)**n) ]
print permute2('abc',2) =, permute2('abc',2)
print len(permute2('13579',3)) =, len(permute2('13579',3))
permute2('abc',2) = ['aa', 'ab', 'ac', 'ba', 'bb', 'bc',
'ca', 'cb', 'cc']
len(permute2('13579',3)) = 125
Charles
--
http://mail.python.org/mailman/listinfo/python-list
Re: Multiple regex match idiom
Hrvoje Niksic wrote: I often have the need to match multiple regexes against a single string, typically a line of input, like this: if (matchobj = re1.match(line)): ... re1 matched; do something with matchobj ... elif (matchobj = re2.match(line)): ... re2 matched; do something with matchobj ... elif (matchobj = re3.match(line)): [snip] There are ways to work around the problem, for example by writing a utility predicate that passes the match object as a side effect, but that feels somewhat non-standard. I'd like to know if there is a Python idiom that I'm missing. What would be the Pythonic way to write the above code? Only just learning Python, but to me this seems better. Completely untested. re_list = [ re1, re2, re3, ... ] for re in re_list: matchob = re.match(line) if matchob: break Of course this only works it the do something is the same for all matches. If not, maybe a function for each case, something like re1 = re.compile() def fn1( s, m ): re2 = def fn2( s, m ): re_list = [ (re1, fn1), (re2, fn2), ... ] for (r,f) in re_list: matchob = r.match(line) if matchob: f( line, matchob ) break f(line,m) Probably better ways than this exist. Charles -- http://mail.python.org/mailman/listinfo/python-list
Re: interesting exercise
[EMAIL PROTECTED] wrote:
On May 9, 1:13 am, Charles Sanders [EMAIL PROTECTED]
wrote:
[snip]
or even this monstrosity ...
def permute2( s, n ):
return [ ''.join([ s[int(i/len(s)**j)%len(s)]
for j in range(n-1,-1,-1)])
for i in range(len(s)**n) ]
print permute2('abc',2) =, permute2('abc',2)
print len(permute2('13579',3)) =, len(permute2('13579',3))
permute2('abc',2) = ['aa', 'ab', 'ac', 'ba', 'bb', 'bc',
'ca', 'cb', 'cc']
len(permute2('13579',3)) = 125
Charles
Could you explain, this one, actually? Don't forget StopIteration.
As Michael said, simple counting in base n with the
string as the digits. No attempt at clarity or efficiency (I
did say it was a monstrosity).
Also, as a python beginner I didn't know about divmod,
and have no idea what you (castironpi) mean by Don't forget
StopIteration.
Charles
--
http://mail.python.org/mailman/listinfo/python-list
Re: 4 quadrant atan
Roel Schroeven wrote: I might be wrong of course, but can't you just use atan2? Only problem is that it returns negative angles for quadrants 3 and 4, but that is easily solved. In Python: from math import atan2, pi, fmod def vectorAngle(x, y): return fmod(atan2(y, x) + 2*pi, 2*pi) No conditionals in sight. Yes, but there will be some conditionals for special cases in the implementation of atan2(). I am pretty sure that in Python (and Perl) atan2() is the C atan2() from the C math library, also often used by Fortran. This is normally written in assembler for the individual architecture by the hardware vendor or compiler writer, is usually heavily optimized, carefully handles all the corner cases, and often carries extra precision to ensure accuracy. If I recall correctly P. J. Plauger's book on the C standard library (The Standard C Library) includes a C version of atan2(), which is probably pretty typical of the algorithms used when there is no hardware support for trigonometric functions. As I remember it, atan2 (and atan) in Plauger's book use a ratio of polynomials to approximate atan(x) over a limited range, (such as -1.0 ... +1.0 or 0.0 ... sqrt(2)-1) and formulae (assuming I have remembered high school trigonometry correctly) such as atan(x) = pi/2 - atan(1/x) and atan(x) = pi/4 - atan((1-x)/(1+x)) [derived from tan(pi/4-x) = (1-x)/(1+x)] to extend the range. There is an obvious trade off between the complexity of the approximating polynomials (and the range they cover with acceptable accuracy), and the number of cases needed to extend the range to all valid inputs. For example, if the polynomials cover -1.0 ... +1.0 (atan() values of -pi/4 to +pi/4) then there are at least 5 cases to consider (atan() ranges -pi ... -3pi/4, -3pi/4 ... -pi/4, -pi/4 ... +pi4, +pi/4 ... +3pi/4, +3pi/4 ... +pi), while if the approximations only cover 0.0 ... sqrt(2)-1 (atan() values from 0.0 to pi/8) then there are at least 16 cases. NaNs and infinities would add additional cases, as may the need to preserve accuracy near zero. Charles -- http://mail.python.org/mailman/listinfo/python-list
Re: Lazy evaluation: overloading the assignment operator?
Diez B. Roggisch wrote: I fail to see where laziness has anything to do with this. In C++, this problem can be remedied with the so called temporary base class idiom. I have seen this referred to as lazy evaluation in C++, so I suspect that Diez and Sturia are using Lazy evaluation in different contexts with different meaning. But this has nothing to do with laziness, which does not reduce the amount of code to execute, but instead defers the point of execution of that code. But that is precisely what Sturia is suggesting, defer (for a few nanoseconds) the evaluation of the multiplications and addition until the assignment occurs. Admittedly a big difference to the lazy evaluation implied by python's yield statement, but still a version of lazy evaluation and (at least sometimes) referred to as such in a C++ context. I am a python newbie (about one month) but I think some of what Sturia wants could be achieved by partially following what is usually done in C++ to achieve what he wants. It would involve a replacement array class (possibly derived from NumPy's arrays) and a proxy class. + Addition, multiplication, etc of arrays and proxy arrays does not return the result array, but returns a proxy which stores the arguments and the operation. + Array indexing of the proxy objects results in the indexing methods of the arguments being called and the operation being carried out and returned. In C++ this is normally very efficient as the operations are all defined inline and expanded by the compiler. + If necessary, define an additional method to evaluate the entire array and return it. I think this would allow code like (if the new array type is XArray) a = Xarray(...) b = Xarray(...) c = Xarray(...) d = Xarray(...) y = a*b+c*d # Returns a proxy object x = y[4] # Computes x = a[4]*b[4] + c[4]*d[4] v = y.eval() # Evaluates all elements, returning Xarray z = ((a+b)*(c+d)).eval() # Also evaluates all elements Whether it would be any faster is doubtful, but it would eliminate the temporaries. Charles -- http://mail.python.org/mailman/listinfo/python-list
Re: scaling
Gabriel Genellina wrote: [snip] if xminvalue: yield 0 elif xmaxvalue: yield top else: yield (x-minvalue)*top/(maxvalue-minvalue) [snip] Personally, I find yield min(top,max(0,(x-minvalue)*top/(maxvalue-minvalue))) or scaled_value = (x-minvalue)*top/(maxvalue-minvalue) yield min(top,max(0,scaled_value)) clearer, but I am aware that others disagree with this. Charles -- http://mail.python.org/mailman/listinfo/python-list
Re: Python not giving free memory back to the os get's me in real problems ...
Grant Edwards wrote: Assuming the python interpreter free()s the memory, my understanding is that on Unixes the memory is returned to the pool used by malloc(), but is not returned to the OS since there isn't a practical way to ensure that the memory at the end of the data segment is not used. http://www.dent.med.uni-muenchen.de/~wmglo/malloc-slides.html Large chunks are allocated via mmap(). The threshold is set through the environment variable MALLOC_MMAP_THRESHOLD_. The maximum number of chunks allocated in this way is limited through MALLOC_MMAP_MAX_. I think these large chunks are returned to the OS when freed. Trimming occurs when the size of the top chunk exceeds MALLOC_TRIM_THRESHOLD_. I think this means that memory for small chunks is returned to the OS if the free space at the top of the heap exceeds the threshold. Asking on the gnu libc mailing list would probably provide a more authoritative answer. I'd wager that they even know what other non-Gnu libc implementations do. Agreed, they would be authoritative for GNU libc. Charles -- http://mail.python.org/mailman/listinfo/python-list
Re: Strange terminal behavior after quitting Tkinter application
Chris wrote: ... Quitting by typing 'sys.exit()' in the interpreter also works fine. Only quitting via the GUI seems to cause the problem. As previously stated, I know nothing about Tkinter, but it definitely looks like there is some cleanup being skipped on a GUI exit that is in fact being run on exit via ^D and sys.exit from the terminal. If I remember correctly, the low level details of how this sort of thing is usually handled on unix like systems is for the application to - Do an ioctl TCGETA system call to get the current settings, and save them somewhere. - Do an ioctl TCSETA system call to change the settings. - Arrange, via signal handlers, atexit, or other means for a cleanup routine to be called on exit. - The cleanup routine does an ioctl TCSETA call with the saved settings. (all this would typically involve C calls) I believe that in your case this last cleanup is being omitted on exit via the GUI. For the moment, including the line 'os.system(stty sane)' before sys.exit() is the solution I'll use. If that is satisfactory, well and good. However, there is a possibility that you may lose some settings that you would prefer to keep. The terminal settings have been trashed, and stty sane has restored a minimally workable set, but some settings may not be what you expect. If you are on a unix like system (eg Linux), and can wrap your Tkinter application in a shell script, you should be able to save/restore the settings outside of Tkinter and python, ie #! /bin/sh saved_settings=`stty -g` i# Note backquotes, save current settings python Tkinter-application.py # Or whatever stty $saved_settings i# Restore settings Doing the save within the Tkinter application, with (for example) os.system(stty -g saved_settings) and os.system(stty `cat saved_settings`) may not work as the setting may have been changed by Tkinter before you get a chance to save them - I just do not know. Just in case, I did a google search. I am not familiar with TKinter, but a couple of articles out there imply that rather than calling sys.exit you should be calling a TkInter routine root.destroy. I am not sure if root is a variable for the main window (ie you call the destroy method on the main window) or if it has some special Tkinter meaning. Presumably this routine cleans things up before calling sys.exit or an equivalent. Charles -- http://mail.python.org/mailman/listinfo/python-list
Re: Compare regular expressions
Thomas Dybdahl Ahle wrote: Hi, I'm writing a program with a large data stream to which modules can connect using regular expressions. Now I'd like to not have to test all expressions every time I get a line, as most of the time, one of them having a match means none of the others can have so. Not what you want, and would be a LOT of work, but if I remember correctly (from university more than 20 years ago) there is an algorithm that could be adapted to return a list of all the regular expressions that match a string. I thing the base algorithms were documented in Aho and Ullman (The Dragon book if I remember correctly). There was one for converting a regular expression into a Non-deterministic Finite-state Automaton, and another for converting the NFA to a Deterministic Finite-state Automaton. The book mentioned that this also handles multiple regular expressions which can be treated as the sub-expressions combined using the or operator. Other, newer books on compiler design would probably contain similar information in their sections on lexical analysis. The algorithm given (if my memory is correct) only handled the case where a true/false match/no-match result was required, but mentioned how to generalise it to return which of several expressions separated by or operators was matched. I think an additional generalisation to return a set of matches rather than one would be relatively straight-forward. I believe these algorithms are the basis of lex/flex and similar utilities, as well as the regular expression handling of many languages. Of course, I would like to emphasise that all this is based on memory of university courses more than 20 years ago, so the details could be wrong. Charles -- http://mail.python.org/mailman/listinfo/python-list
Re: Strange terminal behavior after quitting Tkinter application
Chris wrote: But does anyone know why the Tkinter program is doing this to the terminal in the first place? I don't want to have to tell users of my program that they must recover their terminal's sanity each time after running my program. I don't know about Tkinter, but my guess would be that it has to change the terminal settings to do what it does, and you are quitting without restoring the settings. Is there some Tkinter clean up that you have omitted ? Have you ensured that the clean up runs on both normal exit and abnormal exit (eg ^C) ? For example, the curses library (in C) requires an endwin() call before exit to restore settings. If this is omitted, a stty sane is needed to set the terminal to a semi-sensible default. On Unix and similar systems, signal handlers are normally installed to ensure (as far as possible) that the cleanup occurs if the process is killed. This also applies to vi and similar programs that take total control of the terminal. Charles -- http://mail.python.org/mailman/listinfo/python-list
Re: strange behaviour sys.argv
schnupfy wrote: ok, thanks for the answers. I try to hand over the 3rd part (the long trap) as one cmd argument. I will ask in a shell ng. Thanks again. Cheers Should be as simple as removing the backslashes /root/mk/services.py $HOST $SEVERITY $TRAP should pass TRAP as a single arg Charles -- http://mail.python.org/mailman/listinfo/python-list
Re: strange behaviour sys.argv
Michael Hoffman wrote:
schnupfy wrote:
I am not used to python and I am wondering about this thing:
This is not a Python question. It is a question about how to use bash.
[snip]
Michael is correct, it is a bash thing, nothing to do with python.
bash (and other *nix like shells) generally break arguments on
white space. Quoting (both single and double) overrides this with
(slightly) different rules for different shells.
/root/mk/services.py 192.168.1.101 critical 192.168.1.101
192.168.1.101 SNMPv2-MIB::sysUpTime.0 14:13:02:57.06 SNMPv2-
MIB::snmpTrapOID.0 SNMPv2-SMI::enterprises.789.0.13 SNMPv2-
SMI::enterprises.789.0.2cfCannotTakeover == 1 priority == critical
SNMP-COMMUNITY-MIB::snmpTrapAddress.0 192.168.1.101 SNMP-COMMUNITY-
MIB::snmpTrapCommunity.0 public
Assuming this has been folded and actually is one long
line (which the output confirms), you have passed the python
script seven arguments
'192.168.1.101' (blank seperated)
'critical' (also blank seperated)
a string extending from just after the first double quote to
just before the second, ie starting with '192.168.1.101' and
ending with '789.0.2', with the immediately following (no
white space) unquoted text 'fCannotTakeover' appended
'==' (blank seperated)
'priority
'=='
a string starting with critical, with the quoted string from
'SNMP-COMMUNITY' to 'Community.0 ' (including the blank), the
unquoted string 'public', and the null quoted string all
appended.
TRAP='192.168.1.101 192.168.1.101 SNMPv2-MIB::sysUpTime.0
14:13:02:57.06 SNMPv2-MIB::snmpTrapOID.0 SNMPv2-SMI::enterprises.
789.0.13 SNMPv2-SMI::enterprises.789.0.2cfCannotTakeover == 1
priority == critical SNMP-COMMUNITY-MIB::snmpTrapAddress.0
192.168.1.101 SNMP-COMMUNITY-MIB::snmpTrapCommunity.0 public'
HOST=$(echo $TRAP | awk '{print $1}')
SEVERITY='critical'
/root/mk/services.py $HOST $SEVERITY \$TRAP\
Here, the variables are expanded, and then split into
arguments on white space unless quoted. The backslashes protect
the double quotes so they are treated as normal characters, so
the $TRAP variable is also split into arguments on white space.
Quotes resulting from the substitution of $TRAP are also protected
(ie are treated as ordinary characters).
The result is
'192.168.1.101 (From $HOST)
'critical'(From $SEVERITY)
'192.168.1.101' (Leading '' from \, rest from
$TRAP, blank seperated)
'192.168.1.101' (from $TRAP, blank seperated)
'SNMPv2-MIB::sysUpTime.0'
and so on for the rest of the $TARP string, splitting it at
white space. The last part of $TRAP, 'public', has a double
quote appended from the \.
Python is giving exactly what the shell has given it in both cases.
Charles
--
http://mail.python.org/mailman/listinfo/python-list
Re: Problem with algorithm
Paul Rubin wrote: [snip] def a(n): if n==0: yield '' return for c in s: for r in a(n-1): yield c+r print list(a(3)) Of course, obvious in retrospect, recursion instead of iteration. I have yet to completely wean myself off Fortran style thinking. Charles -- http://mail.python.org/mailman/listinfo/python-list
Re: Problem with algorithm
[EMAIL PROTECTED] wrote: On Apr 12, 10:16�pm, Jia Lu [EMAIL PROTECTED] wrote: Hi all. �I want to create a large list like: ~ Is there any good algorithm to do this? Sure. test = '01' for m in test: for n in test: for o in test: for p in test: print m+n+o+p [snip] Forgive any silly mistakes I have made (I've been teaching myself python for about 1 week) but there is a moderately well known algorithm for this that extends to arbitrary lengths of both the list of alternatives and the length of the required output, and avoids deeply nested loops. I know that it is no better for small and constant output lengths, but for longer lengths or if the output length can vary it should be better. There is a similar algorithm if duplicates are not allowed (ie abcd ... wxyz). My attempt at a python translation of the algorithm: def m_from_n ( v, m ): Print all combinations of m things from v[0] ... v[n-1], duplicates OK. Yields a list. x = [0] * m while True: yield [ v[i] for i in x ] i = m - 1 while i=0 and x[i]==len(v)-1: x[i] = 0 i = i - 1 if i = 0: x[i] = x[i] + 1 else: return for y in m_from_n( xyz, 2 ): print ''.join(y) xx xy xz yx yy yz zx zy zz for y in m_from_n( [0,1], 3 ): print y [0, 0, 0] [0, 0, 1] [0, 1, 0] [0, 1, 1] [1, 0, 0] [1, 0, 1] [1, 1, 0] [1, 1, 1] for y in m_from_n( abcdefghijklmnopqrstuvwxyz, 4 ): print ''.join(y) should more or less do what you want. Charles -- http://mail.python.org/mailman/listinfo/python-list
How to tell easy_install that a package is already installed
I am not sure if this is the right forum to ask this. If it is not, could someone pleas point me to a more appropriate newsgroup. I am attempting to install dap.plugins.netcdf using easy_install on HP-UX 11. As a user, I do not have access to root so have followed the easy_install recommendation to set up a virtual python, as described at http://peak.telecommunity.com/DevCenter/EasyInstall#creating-a-virtual-python and this is mostly working well. One of the dependencies, numpy is already installed, but easy_install does not detect this, and tries to install numpy, which fails with many compilation errors. (numerous pointer type mismatches, undeclared functions like rintf that do not exist on HP, ...) I could try to install numpy myself from the tar.gz file, but this seems pointless since it is already installed. Is there any way to tell easy_install that numpy is installed, while still installing any other dependencies ? Would the best option be to just use the --no-deps option and see if it works ? Charles -- http://mail.python.org/mailman/listinfo/python-list
