plotnine.stat_summary

stat_summary(mapping=None, data=None, **kwargs)

Calculate summary statistics depending on x

{usage}

Parameters

fun_data : str | callable = "mean_cl_boot"

If string, it should be one of:

# Bootstrapped mean, confidence interval
# Arguments:
#     n_samples - No. of samples to draw
#     confidence_interval
#     random_state
"mean_cl_boot"

# Mean, C.I. assuming normal distribution
# Arguments:
#     confidence_interval
"mean_cl_normal"

# Mean, standard deviation * constant
# Arguments:
#     mult - multiplication factor
"mean_sdl"

# Median, outlier quantiles with equal tail areas
# Arguments:
#     confidence_interval
"median_hilow"

# Mean, Standard Errors * constant
# Arguments:
#     mult - multiplication factor
"mean_se"

or any function that takes a array and returns a dataframe with three columns named y, ymin and ymax.

fun_y : callable = None

Any function that takes a array_like and returns a value

fun_ymin : callable = None

Any function that takes an array_like and returns a value

fun_ymax : callable = None

Any function that takes an array_like and returns a value

fun_args : dict = None

Arguments to any of the functions. Provided the names of the arguments of the different functions are in not conflict, the arguments will be assigned to the right functions. If there is a conflict, create a wrapper function that resolves the ambiguity in the argument names.

random_state : int | RandomState = None

Seed or Random number generator to use. If None, then numpy global generator numpy.random is used.

Notes

If any of fun_y, fun_ymin or fun_ymax are provided, the value of fun_data will be ignored.

See Also

geom_pointrange

The default geom for this stat.