there is a hint stated vary the number of estimators between 1 and 60. Can I use
np.linspace(0, 60, num=30), instead of np.logspace(0, 1.8, num = 30)?
np.logspace returns numbers spaced evenly on a logarithmic scale. It is useful only when you are exploring an interval across several orders of magnitude. For the interval (0,60) np.linspace is more adequate.
1 Like
In general when you do hyper-parameter search you can vary the parameters any way you want. As @ArturoAmorQ was saying when the parameter range is “big” (for example from 1e-6
to 1e6
using np.logspace
makes sense).
In this particular case I would guess that picking np.logspace
or np.linspace
will not have too much effect.
1 Like