Styjun

New order #4: World

Can an x86 CPU running in real mode be considered to be basically an 8086 CPU?

What are these boxed doors outside store fronts in New York?

What defenses are there against being summoned by the Gate spell?

My colleague's body is amazing

What typically incentivizes a professor to change jobs to a lower ranking university?

How can I fix this gap between bookcases I made?

Mean and Variance of Continuous Random Variable

Do airline pilots ever risk not hearing communication directed to them specifically, from traffic controllers?

What Brexit solution does the DUP want?

What does "enim et" mean?

Draw simple lines in Inkscape

Patience, young "Padovan"

What do you call a Matrix-like slowdown and camera movement effect?

How is it possible for user's password to be changed after storage was encrypted? (on OS X, Android)

What would happen to a modern skyscraper if it rains micro blackholes?

declaring a variable twice in IIFE

Why don't electron-positron collisions release infinite energy?

Non-Jewish family in an Orthodox Jewish Wedding

What is the white spray-pattern residue inside these Falcon Heavy nozzles?

Why is this code 6.5x slower with optimizations enabled?

How can bays and straits be determined in a procedurally generated map?

Can you lasso down a wizard who is using the Levitate spell?

Why CLRS example on residual networks does not follows its formula?

Applying a recursive function or variable-length loop in to elements of a numpy array

PHP: Delete an element from an arrayUsing global variables in a functionDeleting array elements in JavaScript - delete vs spliceHow do you check if a variable is an array in JavaScript?Get the first element of an arrayLoop through an array in JavaScriptHow do I remove a particular element from an array in JavaScript?How can I add new array elements at the beginning of an array in Javascript?Loop through an array of strings in Bash?Swift for loop: for index, element in array?

.everyoneloves__top-leaderboard:empty,.everyoneloves__mid-leaderboard:empty,.everyoneloves__bot-mid-leaderboard:empty height:90px;width:728px;box-sizing:border-box;

0

I've got a function of the form:

def f(x):

 n = 0

 while x > basecase:
 x = g(x)
 n += 1

 return n

Basically, it applies function g() on x until a condition is met, and then returns how many applications n it took.

I'd like to be able to call this function on numpy arrays, so I can graph f(x) for a large range of x with MatPlotLib. But as the function is now, that doesn't work -- the comparison between x (an array) and basecase (a scalar) is a problem, as is the fact that n is a single integer and not an array.

I know I could just call vectorize, but that's slow -- basically the same as iterating manually. What can I do here? Are there any functional-programming numpy functions that can save me here, or would I have to rewrite things with nditer if I wanted to avoid the overhead of vectorize?

python arrays numpy

share|improve this question

asked Mar 22 at 0:57

a52a52

567

• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  You can rewrite f() and g() using Cython; in any case, you will need to add additional condition to prevent infinity looping. Note: the iterative process you are applying to g(x) is finding root of equation x = g(x). This iterative process will not converge in some cases (this depends on g(x)).
  
  – bubble
  Mar 22 at 1:56
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  nditer is useful only as a stepping stone toward its use in cython. If you must iterate (as opposed to using whole-array numpy operations), work with lists.
  
  – hpaulj
  Mar 22 at 2:50
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  If x is an array, what would a while x>basecase mean? All values of x greater than the target? Any values greater? while is a scalar operation, yes/no. You need to be clear, more to yourself than us, what should happen.
  
  – hpaulj
  Mar 22 at 3:03
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  @bubble g(x) is known, and the process is guaranteed to converge. I just didn't want to include it for sake of simplicity. @hpaulj My goal is for the function to act exactly as if you fed it each element individually, so it tests each element separately and if the test fails, it loops. I'm not sure if something like that is possible (because you would kind of have to interleave python and C loops), which is why I'm asking.
  
  – a52
  Mar 22 at 3:39
  
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  So you aren't solving an array function, but rather a set of independent values. In other words, solving g for a bunch of different starting values.
  
  – hpaulj
  Mar 22 at 4:06
  

  
  
  
  

 | 
show 1 more comment

0

I've got a function of the form:

def f(x):

 n = 0

 while x > basecase:
 x = g(x)
 n += 1

 return n

Basically, it applies function g() on x until a condition is met, and then returns how many applications n it took.

I'd like to be able to call this function on numpy arrays, so I can graph f(x) for a large range of x with MatPlotLib. But as the function is now, that doesn't work -- the comparison between x (an array) and basecase (a scalar) is a problem, as is the fact that n is a single integer and not an array.

I know I could just call vectorize, but that's slow -- basically the same as iterating manually. What can I do here? Are there any functional-programming numpy functions that can save me here, or would I have to rewrite things with nditer if I wanted to avoid the overhead of vectorize?

python arrays numpy

share|improve this question

asked Mar 22 at 0:57

a52a52

567

• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  You can rewrite f() and g() using Cython; in any case, you will need to add additional condition to prevent infinity looping. Note: the iterative process you are applying to g(x) is finding root of equation x = g(x). This iterative process will not converge in some cases (this depends on g(x)).
  
  – bubble
  Mar 22 at 1:56
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  nditer is useful only as a stepping stone toward its use in cython. If you must iterate (as opposed to using whole-array numpy operations), work with lists.
  
  – hpaulj
  Mar 22 at 2:50
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  If x is an array, what would a while x>basecase mean? All values of x greater than the target? Any values greater? while is a scalar operation, yes/no. You need to be clear, more to yourself than us, what should happen.
  
  – hpaulj
  Mar 22 at 3:03
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  @bubble g(x) is known, and the process is guaranteed to converge. I just didn't want to include it for sake of simplicity. @hpaulj My goal is for the function to act exactly as if you fed it each element individually, so it tests each element separately and if the test fails, it loops. I'm not sure if something like that is possible (because you would kind of have to interleave python and C loops), which is why I'm asking.
  
  – a52
  Mar 22 at 3:39
  
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  So you aren't solving an array function, but rather a set of independent values. In other words, solving g for a bunch of different starting values.
  
  – hpaulj
  Mar 22 at 4:06
  

  
  
  
  

 | 
show 1 more comment

I've got a function of the form:

def f(x):

 n = 0

 while x > basecase:
 x = g(x)
 n += 1

 return n

Basically, it applies function g() on x until a condition is met, and then returns how many applications n it took.

I'd like to be able to call this function on numpy arrays, so I can graph f(x) for a large range of x with MatPlotLib. But as the function is now, that doesn't work -- the comparison between x (an array) and basecase (a scalar) is a problem, as is the fact that n is a single integer and not an array.

I know I could just call vectorize, but that's slow -- basically the same as iterating manually. What can I do here? Are there any functional-programming numpy functions that can save me here, or would I have to rewrite things with nditer if I wanted to avoid the overhead of vectorize?

python arrays numpy

share|improve this question

asked Mar 22 at 0:57

a52a52

567

def f(x):

 n = 0

 while x > basecase:
 x = g(x)
 n += 1

 return n

share|improve this question

asked Mar 22 at 0:57

a52a52

567

share|improve this question

asked Mar 22 at 0:57

a52a52

567

asked Mar 22 at 0:57

a52a52

567

asked Mar 22 at 0:57

a52a52

567