I added explanation of how the prediction of error (-0.1333) is done for the fibo sequence.
I removed "recursive" because here in this context (finding say 3 new values in the sequence) the function does not do recursion in the way it's defined in math/computer science (a function calling itself in order to compute its result). You simply meant repeated applications, or iterations, of the main algorithm to the newly formed (longer) sequences.
I removed the Conclusion section because it was unclear in purpose and the expression there at the end is incorrect. If you intended a mathematical formula how the new (prediction with improvement) value in sequence depends on the values in the sequence -- then that is possible, I did it on paper, but its a long derivation and results in a final huge formula involving all previous 4 values with terms like V1*V2*V3^2*V4^2 and similar, which no longer have any special meaning.
As for showing an example that the method produces the correct predictions - it was already done earlier in the Readme.
Of course, feel free to disagree and make changes in my suggestions.
I added explanation of how the prediction of error (-0.1333) is done for the fibo sequence.
I removed "recursive" because here in this context (finding say 3 new values in the sequence) the function does not do recursion in the way it's defined in math/computer science (a function calling itself in order to compute its result). You simply meant repeated applications, or iterations, of the main algorithm to the newly formed (longer) sequences.
I removed the Conclusion section because it was unclear in purpose and the expression there at the end is incorrect. If you intended a mathematical formula how the new (prediction with improvement) value in sequence depends on the values in the sequence -- then that is possible, I did it on paper, but its a long derivation and results in a final huge formula involving all previous 4 values with terms like
V1*V2*V3^2*V4^2and similar, which no longer have any special meaning. As for showing an example that the method produces the correct predictions - it was already done earlier in the Readme.Of course, feel free to disagree and make changes in my suggestions.