- Join over
**1.5M+ people** - Join over
**100K+ communities** - Free
**without limits** - Create
**your own community**

hey channel, i’m trying to add a cumulative median/ stream median to numpy array, but the whole libary is very well abstracted, can anyone point me a direction ? numpy/numpy#14258

add or general accumulation doesn’t support that easily

what's the error in this ?

```
---------------------------------------------------------------------------
NameError Traceback (most recent call last)
<ipython-input-233-6296b41d2324> in <module>
----> 1 class Matrices:
2
3
4 def __init__(self, N=10):
5
<ipython-input-233-6296b41d2324> in Matrices()
27
28
---> 29 self.A = sp.csc_matrix(diags([ [self.h/6]*(self.N), \
30 [self.h/3, *[2*self.h/3]*(self.N-1), self.h/3], \
31 [self.h/6]*(self.N)], [1, 0, -1] ), dtype=np.float32)
NameError: name 'self' is not defined
```

I want to compute value at t=1

starting from t=0

but step is too slow

but what is atol

rtol?

also I have two extra time independent initial conditions

and each time I want the 0, -1 entry to be according to init conditins ie., 0, 0

and that time step should be h**2 always where h=1/N and N is lenth of y

y = initial value

?

how to dump numpy to pickle?

I have an object class of sparse matrices and lambdas etc . How do I pickle that thing

?

pickle is not important at the moment

I have written a program to solve the heat eqn using RK45 of scipy using the finite element method

but increasing the number of basis functions the error increases

its counterintuitive

plz help

I m using hat functions homogenous equation

@XtremeGood asking precise and focused questions containing code snippets is very likely to improve the quality of the feed-back you get. I would suggest that you read the following resources: https://opensource.guide/how-to-contribute/#communicating-effectively, https://stackoverflow.com/help/how-to-ask and https://stackoverflow.com/help/minimal-reproducible-example.

see the rk45 class there

why is this code taking too much time?

I 'm using sparse matrices.

csr to be more specific