In the description of the abs() function there is no mention that this can create unpredicted results on float.
I have proven this by writing a small sketch where it works fine, but when using abs() in my bigger project it fails when processing the exact same number (-0.0027).
Please update the description for the abs() function and tell us to use fabs instead.
thats pretty much standard C "features" but I agree it should be updated
Though I dont know who has powers over the reference pages, and if its anything like the playground pages its simpler to answer questions than learn that stupid arse wiki markup that no one uses outside of wikipedia