"values.")
scalar_names = ', '.join(df["name"].unique())
confidence_level_str = props[
"confidence_level"] if "confidence_level" in props else "none"
if confidence_level_str == "none":
df = pd.pivot_table(df,
values="value",
columns=iso_itervars,
index=xaxis_itervar)
errors_df = None
else:
confidence_level = float(confidence_level_str[:-1]) / 100
conf_int = lambda values: utils.confidence_interval(
confidence_level, values) if len(values) > 1 else math.nan
pivoted = pd.pivot_table(df,
values="value",
columns=iso_itervars,
index=xaxis_itervar if xaxis_itervar else "name",
aggfunc=[np.mean, conf_int],
dropna=False)
df = pivoted["mean"]
errors_df = pivoted["<lambda>"]
legend_cols, _ = utils.extract_label_columns(df)
p = plot if chart.is_native_chart() else plt
try:
def conf_intv(values):
return utils.confidence_interval(confidence_level, values)