Does anyone on here know what I can do to the function in the

File: helpmesci2.png (23KB, 735x247px) Image search: [Google]

23KB, 735x247px

Alternatively (it's the same problem), I need to find a way to make the graph in orange here more like the graph in blue, black, or as a ridiculously extreme case, red.

Please /sci/ I need your help ;_;

Is there an exponential fit equivalent to the polynomial fit? Because the polynomial fit doesn't work for shit in this case and this appears to be all exponentials with constants somewhere I don't know where and what amount.

I think the black part looks like a logistic sigmoid

>>9090621
Thanks a lot!
>logistic and sigmoid
What's the difference between both? This seems to be close though, so very appreciated.

Oh, and is there a way to make the logistic/sigmoid function reach the point 1 when x=0, and 0 when x=1 (instead of it going into infinity without ever reaching 0 nor 1)?

File: soclose.png (43KB, 1047x463px) Image search: [Google]

43KB, 1047x463px

I'm so close! How to I fix the last segment though?

>>9090351
Find a separate equation that precisely matches the delta of your desired graph and the displayed graph, then add the two together. The result should be precisely what you want.

Bump for this
>>9090673
>>9090653

Guys, please, I need help with this ;_; >>9090673

>>9090673
>>9091754
Why not just use some interpolation routine?

>>9090351
Step function

>>9090351

Use a sigmoid. Steeper sigmoids use a higher value in the exponent. Try playing with this:

[math] \frac{1}{1 + e^{-x}} [/math]

Just raise the value of [math] x [/math] until you are satisfied.

>>9091822
I already tried that, but it leaves me with this >>9090673, so I need to find a way to "accelerate" the Xs or "attenuate" the Ys towards the end.

>>9091786
How is that different from polynomial fit?

>>9091830

Where are your measurements coming from ? Nature ? Simulation ? Back of a cereal box ?

>>9091840
It's difficult to explain. Why?
Values go from (1,0) to (0,1) though.

>>9091845

Certain data lends certain fits. Population dynamics appreciates sigmoidals and logistic curves, where as studying solar activity might be more sinusoidal.