AdaBoost Alpha

Compute the weight $\alpha \backslash alpha$ AdaBoost gives a weak learner from its weighted error rate $\epsilon \backslash varepsilon$.

Implement solve(error).

Returns. Return a float. The error must lie strictly between 0 and 1 (otherwise ValueError). gives a positive weight, $\epsilon =0.5\backslash varepsilon=0.5$ gives $00$, and $\epsilon >0.5\backslash varepsilon>0.5$ gives a negative weight.

Examples

Example 1

Input:

solve(0.2)

Output:

0.693147

Example 2

Input:

solve(0.5)

Output:

0.0

Example 3

Input:

solve(0.0)

Output: Raises ValueError.