gini can be negative by your code

[arthur-yh]

Gini index can be evaluated to be negative by your code, "The Gini coefficient is double of the area between the Lorenz Curve and the 45◦ diagonal line", so in your code metrics.py::line56, it should be like raw = df.apply(lambda x: abs(2 * x.sum() / df.shape[0] - 1.))
while I push a mr to you which shows a example of gini can be negative and it pass the unitest .
this case shows a bad model which predict high-life-value users's scores lowest, so it's gini is negative since it's line is under the Lorenz Curve.

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.