Using numpy array with Org Babel

Question

I want to pass in 2 arrays and have python process them. The babel block works when I put the arrays directly in but if I try to pass them in via headers, they don't appear as arrays to numpy. My example follows for the failure is below.

#+NAME: p-list
[0.2757333815097, -0.0010721227154135, 0.0002933239156845]
#+NAME: q-list
[0.27571909, -0.00107294,  0.00029381]

#+begin_src python :var p=p-list :var q=q-list :results output listings :exports results
import sys
sys.path.append('./')
import numpy as np
import simple
#parr = np.array([0.2757333815097, -0.0010721227154135, 0.0002933239156845])
#qarr = np.array([0.27571909, -0.00107294,  0.00029381])
parr = np.array(p)
qarr = np.array(q)
print(simple.doit(parr,qarr))
#+end_src

#+RESULTS:

The doit function just calls numpy.dot()

NickD · Accepted Answer · 2023-07-20 11:31:47Z

Try this:

#+NAME: p-list
#+begin_src python :results value
return [0.2757333815097, -0.0010721227154135, 0.0002933239156845]
#+end_src

#+NAME: q-list
#+begin_src python :results value
return [0.27571909, -0.00107294,  0.00029381]
#+end_src

#+begin_src python :var p=p-list() :var q=q-list() :results output listings :exports results
import sys
sys.path.append('./')
import numpy as np

#parr = np.array([0.2757333815097, -0.0010721227154135, 0.0002933239156845])
#qarr = np.array([0.27571909, -0.00107294,  0.00029381])
parr = np.array(p)
qarr = np.array(q)
print(np.dot(parr,qarr))
#+end_src

#+RESULTS:
: 0.07602619353732326

The variables p and q in the last source block are assigned the values obtained by evaluating the first two code blocks - note the parens.

What you are doing is passing strings to the last block. Consider this:

#+name: foo
[3, 4, 5]

#+begin_src python :var x=foo :results output
print(x)
print(type(x))
#+end_src

#+RESULTS:
: [3, 4, 5]
: 
: <class 'str'>                        <==== It's a string!

EDIT: To reply to the OP's question in a comment: Yes, it is possible to use the simpler string representation; you just need to convert the string to a list inside the Python program.

A way to do that is to use the Python eval function in the program to convert the string to a Python list. Here's a modification of the last example to illustrate:

#+name: foo
[3, 4, 5]

#+begin_src python :var x=foo :results output
print(x)
print(type(x))
x = eval(x)
print(x)
print(type(x))
#+end_src


#+RESULTS:
: [3, 4, 5]
: 
: <class 'str'>
: [3, 4, 5]
: <class 'list'>

Afthe the eval, x is now a list, so you can convert it to a numpy array. Applying this idea to the original program gives:

#+NAME: p-list-as-string
[0.2757333815097, -0.0010721227154135, 0.0002933239156845]

#+NAME: q-list-as-string
[0.27571909, -0.00107294,  0.00029381]

#+begin_src python :var p=p-list-as-string :var q=q-list-as-string :results output listings :exports results
  import sys
  sys.path.append('./')
  import numpy as np

  parr = np.array(eval(p))
  qarr = np.array(eval(q))
  print(np.dot(parr,qarr))
#+end_src

#+RESULTS:
: 0.07602619353732326

Your answer works. Thanks. Is there a way I can make the vectors appear in the document as normal text vectors. Putting them inside python source blocks isn't how I expect the document to look. Can this work with LaTeX verbatim also? If not, I will still use your solution. — user1625344
– user1625344, Commented Jul 20, 2023 at 2:06
See edit for the first part. I don't know what you mean by 'Can this work with LaTeX verbatim also?` but it sounds like a different question, so you should probably create a new question for it. — NickD
– NickD, Commented Jul 20, 2023 at 10:50

Stack Exchange Network

Using numpy array with Org Babel

1 Answer 1

Your Answer

Hot Network Questions

Using numpy array with Org Babel

1 Answer 1

Your Answer

Sign up or log in

Post as a guest

Related

Hot Network Questions