{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import hvplot.pandas  # noqa"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`errorbars` are most helpful when overlayed with other plot types. To do this we'll use the `*` operator."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from bokeh.sampledata import perceptions\n",
    "\n",
    "data = perceptions.numberly.describe().transpose().sort_values('mean')\n",
    "data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we'll use just one column to set the size of the errorbars around 'mean'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.hvplot.scatter(y='mean', ylabel='amount') * data.hvplot.errorbars(y='mean', yerr1='std')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then we'll use two columns. Remember that these are not the absolute placements of the ends of the bars, but the distance from the center."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "data['yerr1'] = data['mean'] - data['min']\n",
    "data['yerr2'] = data['max'] - data['mean']\n",
    "data.hvplot.scatter(y='mean', ylabel='amount') * data.hvplot.errorbars(y='mean', yerr1='yerr1', yerr2='yerr2')"
   ]
  }
 ],
 "metadata": {
  "language_info": {
   "name": "python",
   "pygments_lexer": "ipython3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}