Reverse sigmoid function in python

Question

How can I create a reverse sigmoid function? I have created the following one but am not getting the desired output.

def sigmoid(x):
  return - (1 / (1 + math.exp(0.5*x)))

x= np.linspace(0,5,100)
y = np.vectorize(sigmoid)
plt.plot(x,y(x))

Please notice the x range should be between 0 -5

UPDATE :

I am looking for the following shape

Wait, so do you want the reverse function of `sigmoid(x)`? I'm not too familiary with `np`, but what if you plotted `(y(x), x)`? — Ryan Zhang, Jul 04 '22 at 13:26
. I want to reverse a sigmoid .. thats my approch ( which seems not right I have added - minus to do that ) — AnonymousScientificUser, Jul 04 '22 at 13:27
Use `-x`? Hard to give an answer without knowing what "reverse" means. — xdurch0, Jul 04 '22 at 13:30
Like, you want some `g(x)` such that `sigmoid(g(x)) == x` holds true for all x? — Ryan Zhang, Jul 04 '22 at 13:31
What do you mean the x range should be between 0 and 5? The domain of a sigmoid is the whole real line, so if you reverse it, it will still be the whole real line. — nonDucor, Jul 04 '22 at 13:41

score 0 · Answer 1 · answered Jul 04 '22 at 13:34

0

If you just want to invert the x axis, you want a function such that g(x) = sigmoid(-x), your definition should be:

def rev_sigmoid(x):
  return (1 / (1 + math.exp(0.5*x)))

x= np.linspace(-10,10,100)
y = np.vectorize(rev_sigmoid)
plt.plot(x,y(x))

answered Jul 04 '22 at 13:34

nonDucor

I have managed it almost . thanks though – AnonymousScientificUser Jul 04 '22 at 13:38
@AnonymousScientificUser: what do you mean with "but in the x constain between 0-5"? – nonDucor Jul 04 '22 at 13:40

1 Answers1