Applying box count method to array of stock data

[laclimber]

I want to apply the box count method to the last 20 years of stock price data. I have an array with shape 4946 X 2 in numpy form. But when I apply the box count function to it, I don't get any reasonable results. Just this:

[jgostick]

Is the image type boolean? And what does it look like?

[laclimber]

the array is type numpy.ndarray . and looks like this:
array([[0.00000000e+00, 6.77971734e+00],
[1.00000000e+00, 6.81237810e+00],
[2.00000000e+00, 6.81189395e+00],
...,
[4.94300000e+03, 8.47492432e+00],
[4.94400000e+03, 8.47193010e+00],
[4.94500000e+03, 8.46930042e+00]])

I put the price in log form

[jgostick]

Im sorry I didn't realize you were doing time data. The function was made for boolean images, and is definitely not suitable for your case.

[laclimber]

i was able to convert into a boolean image (from the picture of the chart) and got the box count method to work. But how do i find the exact fractal dimension?

[laclimber]

[jgostick]

It is the slope of the line on the left, which is what is plotted on the right. Seems to be about 1, so not very fractal.