Currently, outliers are defined as being +/- 2 std. dev. from the mean. For the graph, this would be displayed as a background ribbon (in R we call these ribbon plots, but apparently with matplotlib it's fill_between).
There would also be a new section of the report listing the furthest outliers - perhaps a text section with how many outliers there are, what the median absolute std dev of these outliers are, any other useful stats on the outliers, then a list of highest outliers and lowest outliers, sorted by absolute std. dev. (max 20 each side?),
Currently, outliers are defined as being +/- 2 std. dev. from the mean. For the graph, this would be displayed as a background ribbon (in R we call these ribbon plots, but apparently with matplotlib it's fill_between).
There would also be a new section of the report listing the furthest outliers - perhaps a text section with how many outliers there are, what the median absolute std dev of these outliers are, any other useful stats on the outliers, then a list of highest outliers and lowest outliers, sorted by absolute std. dev. (max 20 each side?),