We Keep Coding

Is there an elegant way to check if index can be requested in a numpy array?

I am looking for an elegant way to check if a given index is inside a numpy array (for example for BFS algorithms on a grid).
The following code does what I want:
import numpy as np
def isValid(np_shape: tuple, index: tuple):
if min(index) < 0:
return False
for ind,sh in zip(index,np_shape):
if ind >= sh:
return False
return True
arr = np.zeros((3,5))
print(isValid(arr.shape,(0,0))) # True
print(isValid(arr.shape,(2,4))) # True
print(isValid(arr.shape,(4,4))) # False
But I'd prefer something build-in or more elegant than writing my own function including python for-loops (yikes)

You can try:
def isValid(np_shape: tuple, index: tuple):
index = np.array(index)
return (index >= 0).all() and (index < arr.shape).all()
arr = np.zeros((3,5))
print(isValid(arr.shape,(0,0))) # True
print(isValid(arr.shape,(2,4))) # True
print(isValid(arr.shape,(4,4))) # False

I have benchmarked the answers quite a bit, and come to the conclusion that actually the explicit for loop as provided in my code performs best.
Dmitri's solution is wrong for several reasons (tuple1 < tuple2 just compares the first value; ideas like np.all(ni < sh for ind,sh in zip(index,np_shape)) fail as the input to all returns a generator, not a list etc).
#mozway's solution is correct, but all the casts make it a lot slower. Also it always needs to consider all numbers for casting, while an explicit loop can stop earlier, I suppose.
Here is my benchmark (Method 0 is #mozway's solution, Method 1 is my solution):

`numpy.nanpercentile` is extremely slow

numpy.nanpercentile is extremely slow.
So, I wanted to use cupy.nanpercentile; but there is not cupy.nanpercentile implemented yet.
Do someone have solution for it?

I also had a problem with np.nanpercentile being very slow for my datasets. I found a wokraround that lets you use the standard np.percentile. And it can also be applied to many other libs.
This one should solve your problem. And it also works alot faster than np.nanpercentile:
arr = np.array([[np.nan,2,3,1,2,3],
[np.nan,np.nan,1,3,2,1],
[4,5,6,7,np.nan,9]])
mask = (arr >= np.nanmin(arr)).astype(int)
count = mask.sum(axis=1)
groups = np.unique(count)
groups = groups[groups > 0]
p90 = np.zeros((arr.shape[0]))
for g in range(len(groups)):
pos = np.where (count == groups[g])
values = arr[pos]
values = np.nan_to_num (values, nan=(np.nanmin(arr)-1))
values = np.sort (values, axis=1)
values = values[:,-groups[g]:]
p90[pos] = np.percentile (values, 90, axis=1)
So instead of taking the percentile with the nans, it sorts the rows by the amount of valid data, and takes the percentile of those rows separated. Then adds everything back together. This also works for 3D-arrays, just add y_pos and x_pos instead of pos. And watch out for what axis you are calculating over.

def testset_gen(num):
init=[]
for i in range (num):
a=random.randint(65,122) # Dummy name
b=random.randint(1,100) # Dummy value: 11~100 and 10% of nan
if b<11:
b=np.nan # 10% = nan
init.append([a,b])
return np.array(init)
np_testset=testset_gen(30000000) # 468,751KB
def f1_np (arr, num):
return np.percentile (arr[:,1], num)
# 55.0, 0.523902416229248 sec
print (f1_np(np_testset[:,1], 50))
def cupy_nanpercentile (arr, num):
return len(cp.where(arr > num)[0]) / (len(arr) - cp.sum(cp.isnan(arr))) * 100
# 55.548758317136446, 0.3640251159667969 sec
# 43% faster
# If You need same result, use int(). But You lose saved time.
print (cupy_nanpercentile(cp_testset[:,1], 50))
I can't imagine How test result takes few days. With my computer, It seems 1 Trillion line of data or more. Because of this, I can't reproduce same problem due to lack of resource.

Here's an implementation with numba. After it's been compiled it is more than 7x faster than the numpy version.
Right now it is set up to take the percentile along the first axis, however it could be changed easily.
#numba.jit(nopython=True, cache=True)
def nan_percentile_axis0(arr, percentiles):
"""Faster implementation of np.nanpercentile
This implementation always takes the percentile along axis 0.
Uses numba to speed up the calculation by more than 7x.
Function is equivalent to np.nanpercentile(arr, <percentiles>, axis=0)
Params:
arr (np.array): Array to calculate percentiles for
percentiles (np.array): 1D array of percentiles to calculate
Returns:
(np.array) Array with first dimension corresponding to
values as passed in percentiles
"""
shape = arr.shape
arr = arr.reshape((arr.shape[0], -1))
out = np.empty((len(percentiles), arr.shape[1]))
for i in range(arr.shape[1]):
out[:,i] = np.nanpercentile(arr[:,i], percentiles)
shape = (out.shape[0], *shape[1:])
return out.reshape(shape)

Processing big list using python

So I'm trying to solve a challenge and have come across a dead end. My solution works when the list is small or medium but when it is over 50000. It just "time out"
a = int(input().strip())
b = list(map(int,input().split()))
result = []
flag = []
for i in range(len(b)):
temp = a - b[i]
if(temp >=0 and temp in flag):
if(temp<b[i]):
result.append((temp,b[i]))
else:
result.append((b[i],temp))
flag.remove(temp)
else:
flag.append(b[i])
result.sort()
for i in result:
print(i[0],i[1])
Where
a = 10
and b = [ 2, 4 ,6 ,8, 5 ]
Solution sum any two element in b which matches a
**Edit: ** Updated full code

flag is a list, of potentially the same order of magnitude as b. So, when you do temp in flag that's a linear search: it has to check every value in flag to see if that value is == temp. So, that's 50000 comparisons. And you're doing that once per loop in a linear walk over b. So, your total time is quadratic: 50,000 * 50,000 = 2,500,000,000. (And flag.remove is also linear time.)
If you replace flag with a set, you can test it for membership (and remove from it) in constant time. So your total time drops from quadratic to linear, or 50,000 steps, which is a lot faster than 2 billion:
flagset = set(flag)
for i in range(len(b)):
temp = a - b[i]
if(temp >=0 and temp in flagset):
if(temp<b[i]):
result.append((temp,b[i]))
else:
result.append((b[i],temp))
flagset.remove(temp)
else:
flagset.add(b[i])
flag = list(flagset)
If flag needs to retain duplicate values, then it's a multiset, not a set, which means you can implement with Counter:
flagset = collections.Counter(flag)
for i in range(len(b)):
temp = a - b[i]
if(temp >=0 and flagset[temp]):
if(temp<b[i]):
result.append((temp,b[i]))
else:
result.append((b[i],temp))
flagset[temp] -= 1
else:
flagset[temp] += 1
flag = list(flagset.elements())
In your edited code, you’ve got another list that’s potentially of the same size, result, and you’re sorting that list every time through the loop.
Sorting takes log-linear time. Since you do it up to 50,000 times, that’s around log(50;000) * 50,000 * 50,000, or around 30 billion steps.
If you needed to keep result in order throughout the operation, you’d want to use a logarithmic data structure, like a binary search tree or a skiplist, so you could insert a new element in the right place in logarithmic time, which would mean just 800.000 steps.
But you don’t need it in order until the end. So, much more simply, just move the result.sort out of the loop and do it at the end.

Evaluating a function using numpy

What is the significance of the return part when evaluating functions? Why is this necessary?

Your assumption is right: dfdx[0] is indeed the first value in that array, so according to your code that would correspond to evaluating the derivative at x=-1.0.
To know the correct index where x is equal to 0, you will have to look for it in the x array.
One way to find this is the following, where we find the index of the value where |x-0| is minimal (so essentially where x=0 but float arithmetic requires taking some precautions) using argmin :
index0 = np.argmin(np.abs(x-0))
And we then get what we want, dfdx at the index where x is 0 :
print dfdx[index0]
An other but less robust way regarding float arithmetic trickery is to do the following:
# we make a boolean array that is True where x is zero and False everywhere else
bool_array = (x==0)
# Numpy alows to use a boolean array as a way to index an array
# Doing so will get you the all the values of dfdx where bool_array is True
# In our case that will hopefully give us dfdx where x=0
print dfdx[bool_array]
# same thing as oneliner
print dfdx[x==0]

You give the answer. x[0] is -1.0, and you want the value at the middle of the array.`np.linspace is the good function to build such series of values :
def f1(x):
g = np.sin(math.pi*np.exp(-x))
return g
n = 1001 # odd !
x=linspace(-1,1,n) #x[n//2] is 0
f1x=f1(x)
df1=np.diff(f1(x),1)
dx=np.diff(x)
df1dx = - math.pi*np.exp(-x)*np.cos(math.pi*np.exp(-x))[:-1] # to discard last element
# In [3]: np.allclose(df1/dx,df1dx,atol=dx[0])
# Out[3]: True
As an other tip, numpy arrays are more efficiently and readably used without loops.

Efficient ways to iterate over several numpy arrays and process current and previous elements?

I've read a lot about different techniques for iterating over numpy arrays recently and it seems that consensus is not to iterate at all (for instance, see a comment here). There are several similar questions on SO, but my case is a bit different as I have to combine "iterating" (or not iterating) and accessing previous values.
Let's say there are N (N is small, usually 4, might be up to 7) 1-D numpy arrays of float128 in a list X, all arrays are of the same size. To give you a little insight, these are data from PDE integration, each array stands for one function, and I would like to apply a Poincare section. Unfortunately, the algorithm should be both memory- and time-efficient since these arrays are sometimes ~1Gb each, and there are only 4Gb of RAM on board (I've just learnt about memmap'ing of numpy arrays and now consider using them instead of regular ones).
One of these arrays is used for "filtering" the others, so I start with secaxis = X.pop(idx). Now I have to locate pairs of indices where (secaxis[i-1] > 0 and secaxis[i] < 0) or (secaxis[i-1] < 0 and secaxis[i] > 0) and then apply simple algebraic transformations to remaining arrays, X (and save results). Worth mentioning, data shouldn't be wasted during this operation.
There are multiple ways for doing that, but none of them seem efficient (and elegant enough) to me. One is a C-like approach, where you just iterate in a for-loop:
import array # better than lists
res = [ array.array('d') for _ in X ]
for i in xrange(1,secaxis.size):
if condition: # see above
co = -secaxis[i-1]/secaxis[i]
for j in xrange(N):
res[j].append( (X[j][i-1] + co*X[j][i])/(1+co) )
This is clearly very inefficient and besides not a Pythonic way.
Another way is to use numpy.nditer, but I haven't figured out yet how one accesses the previous value, though it allows iterating over several arrays at once:
# without secaxis = X.pop(idx)
it = numpy.nditer(X)
for vec in it:
# vec[idx] is current value, how do you get the previous (or next) one?
Third possibility is to first find sought indices with efficient numpy slices, and then use them for bulk multiplication/addition. I prefer this one for now:
res = []
inds, = numpy.where((secaxis[:-1] < 0) * (secaxis[1:] > 0) +
(secaxis[:-1] > 0) * (secaxis[1:] < 0))
coefs = -secaxis[inds] / secaxis[inds+1] # array of coefficients
for f in X: # loop is done only N-1 times, that is, 3 to 6
res.append( (f[inds] + coefs*f[inds+1]) / (1+coefs) )
But this is seemingly done in 7 + 2*(N - 1) passes, moreover, I'm not sure about secaxis[inds] type of addressing (it is not slicing and generally it has to find all elements by indices just like in the first method, doesn't it?).
Finally, I've also tried using itertools and it resulted in monstrous and obscure structures, which might stem from the fact that I'm not very familiar with functional programming:
def filt(x):
return (x[0] < 0 and x[1] > 0) or (x[0] > 0 and x[1] < 0)
import array
from itertools import izip, tee, ifilter
res = [ array.array('d') for _ in X ]
iters = [iter(x) for x in X] # N-1 iterators in a list
prev, curr = tee(izip(*iters)) # 2 similar iterators, each of which
# consists of N-1 iterators
next(curr, None) # one of them is now for current value
seciter = tee(iter(secaxis))
next(seciter[1], None)
for x in ifilter(filt, izip(seciter[0], seciter[1], prev, curr)):
co = - x[0]/x[1]
for r, p, c in zip(res, x[2], x[3]):
r.append( (p+co*c) / (1+co) )
Not only this looks very ugly, it also takes an awful lot of time to complete.
So, I have following questions:
Of all these methods is the third one indeed the best? If so, what can be done to impove the last one?
Are there any other, better ones yet?
Out of sheer curiosity, is there a way to solve the problem using nditer?
Finally, will I be better off using memmap versions of numpy arrays, or will it probably slow things down a lot? Maybe I should only load secaxis array into RAM, keep others on disk and use third method?
(bonus question) List of equal in length 1-D numpy arrays comes from loading N .npy files whose sizes aren't known beforehand (but N is). Would it be more efficient to read one array, then allocate memory for one 2-D numpy array (slight memory overhead here) and read remaining into that 2-D array?

The numpy.where() version is fast enough, you can speedup it a little by method3(). If the > condition can change to >=, you can also use method4().
import numpy as np
a = np.random.randn(100000)
def method1(a):
idx = []
for i in range(1, len(a)):
if (a[i-1] > 0 and a[i] < 0) or (a[i-1] < 0 and a[i] > 0):
idx.append(i)
return idx
def method2(a):
inds, = np.where((a[:-1] < 0) * (a[1:] > 0) +
(a[:-1] > 0) * (a[1:] < 0))
return inds + 1
def method3(a):
m = a < 0
p = a > 0
return np.where((m[:-1] & p[1:]) | (p[:-1] & m[1:]))[0] + 1
def method4(a):
return np.where(np.diff(a >= 0))[0] + 1
assert np.allclose(method1(a), method2(a))
assert np.allclose(method2(a), method3(a))
assert np.allclose(method3(a), method4(a))
%timeit method1(a)
%timeit method2(a)
%timeit method3(a)
%timeit method4(a)
the %timeit result:
1 loop, best of 3: 294 ms per loop
1000 loops, best of 3: 1.52 ms per loop
1000 loops, best of 3: 1.38 ms per loop
1000 loops, best of 3: 1.39 ms per loop

I'll need to read your post in more detail, but will start with some general observations (from previous iteration questions).
There isn't an efficient way of iterating over arrays in Python, though there are things that slow things down. I like to distinguish between the iteration mechanism (nditer, for x in A:) and the action (alist.append(...), x[i+1] += 1). The big time consumer is usually the action, done many times, not the iteration mechanism itself.
Letting numpy do the iteration in compiled code is the fastest.
xdiff = x[1:] - x[:-1]
is much faster than
xdiff = np.zeros(x.shape[0]-1)
for i in range(x.shape[0]:
xdiff[i] = x[i+1] - x[i]
The np.nditer isn't any faster.
nditer is recommended as a general iteration tool in compiled code. But its main value lies in handling broadcasting and coordinating the iteration over several arrays (input/output). And you need to use buffering and c like code to get the best speed from nditer (I'll look up a recent SO question).
https://stackoverflow.com/a/39058906/901925
Don't use nditer without studying the relevant iteration tutorial page (the one that ends with a cython example).
=========================
Just judging from experience, this approach will be fastest. Yes it's going to iterate over secaxis a number of times, but those are all done in compiled code, and will be much faster than any iteration in Python. And the for f in X: iteration is just a few times.
res = []
inds, = numpy.where((secaxis[:-1] < 0) * (secaxis[1:] > 0) +
(secaxis[:-1] > 0) * (secaxis[1:] < 0))
coefs = -secaxis[inds] / secaxis[inds+1] # array of coefficients
for f in X:
res.append( (f[inds] + coefs*f[inds+1]) / (1+coefs) )
#HYRY has explored alternatives for making the where step faster. But as you can see the differences aren't that big. Other possible tweaks
inds1 = inds+1
coefs = -secaxis[inds] / secaxis[inds1]
coefs1 = coefs+1
for f in X:
res.append(( f[inds] + coefs*f[inds1]) / coefs1)
If X was an array, res could be an array as well.
res = (X[:,inds] + coefs*X[:,inds1])/coefs1
But for small N I suspect the list res is just as good. Don't need to make the arrays any bigger than necessary. The tweaks are minor, just trying to avoid recalculating things.
=================
This use of np.where is just np.nonzero. That actually makes two passes of the array, once with np.count_nonzero to determine how many values it will return, and create the return structure (list of arrays of now known length). And a second loop to fill in those indices. So multiple iterations are fine if it keeps action simple.