actsnclass.make_metrics_plots

actsnclass.make_metrics_plots(user_input)

Generate metric plots.

Parameters:

• -m (list) – List of paths to the metrics files. One path for each learning strategy we wish to plot. This must follow the same order as give in -s.

• -o (str) – Path to output file where the plot will be stored.

• -s (list) – List of keywords describing the learning strategies to be plotted. Order must be the same as provided in -m. Options are [‘canonical’, ‘rand_sampling’, ‘unc_sampling’]

Examples

Use it directly from the command line. For example, if you wish to make a metric plot for the random sampling and uncertainty sampling together, do:

>>> make_metrics_plots.py -m <path to rand sampling metrics> <path to unc sampling metrics>
>>>     -o <path to output plot file> -s RandomSampling UncSampling