Firstly, amazing work on the crate!
I was wondering whether you would consider having more informative errors for when the search fails. For example, in the case of non-convergence, returning what the last estimate or last bracket was.
I realize that the different algorithms work differently and thus may store different underlying information, but I thought this could be particularly useful if one wanted to chain two root finding methods in case the first one fails.
Firstly, amazing work on the crate!
I was wondering whether you would consider having more informative errors for when the search fails. For example, in the case of non-convergence, returning what the last estimate or last bracket was.
I realize that the different algorithms work differently and thus may store different underlying information, but I thought this could be particularly useful if one wanted to chain two root finding methods in case the first one fails.