{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Lesson 46: Plotting with Matplotlib and Seaborn\n", "\n", "
" ] }, { "cell_type": "code", "execution_count": 1, "metadata": { "tags": [] }, "outputs": [], "source": [ "import pandas as pd\n", "\n", "import matplotlib.pyplot as plt\n", "import seaborn as sns\n", "\n", "import seaborn.objects as so\n", "\n", "# Magic function to make matplotlib inline; other style specs must come AFTER\n", "%matplotlib inline\n", "\n", "# This enables SVG graphics inline\n", "%config InlineBackend.figure_format = 'svg'" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "
\n", "\n", "In this lesson, we will learn how to use Matplotlib and Seaborn by making many of the same plots as in our [intro lesson on Bokeh](l18_plotting.ipynb) and on [high-level plotting](l19_high_level_plotting.ipynb). To start with, we will use the Glasgow face matching data." ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "tags": [] }, "outputs": [], "source": [ "df_gfmt = pd.read_csv('data/gfmt_sleep.csv', na_values='*')\n", "df_gfmt['insomnia'] = df_gfmt['sci'] <= 16" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## A Matplotlib scatter plot\n", "\n", "A graphic composed with Matplotlib consists of a **figure** (the graphic itself) and one or more **axes**, each an individual plot. Figures and axes are easily created using the `matplotlib.pyplot.subplots()` function, which returns a `Figure` object and a collection (or a single) `Axes` objects. The `matplotlib.pyplot` submodule is traditionally imported as `plt`, as we have done in this notebook. To get a single plot with axes that are 4 inches by 4 inches, we use\n", "\n", " fig, ax = plt.subplots(figsize=(4, 4))\n", " \n", "The `ax` object, then, has many methods. Most importantly, the `ax.plot()` method allows for making line and scatter plots. To make a scatter plot, we use the `marker='.'` and `linestyle=''` kwargs. The plot below is an example, with the various methods of the `ax` object being self-explanatory." ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "tags": [] }, "outputs": [ { "data": { "image/svg+xml": [ "\n", "\n", "\n", " \n", " \n", " \n", " \n", " 2023-06-29T14:39:04.193879\n", " image/svg+xml\n", " \n", " \n", " Matplotlib v3.7.1, https://matplotlib.org/\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "\n" ], "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "fig, ax = plt.subplots(figsize=(4, 4))\n", "ax.set_xlabel(\"confidence when correct\")\n", "ax.set_ylabel(\"confidence when incorrect\")\n", "ax.set_xlim(-2.5, 102.5)\n", "ax.set_ylim(-2.5, 102.5)\n", "ax.grid(True)\n", "\n", "ax.plot(\n", " \"confidence when correct\",\n", " \"confidence when incorrect\",\n", " data=df_gfmt,\n", " marker=\".\",\n", " linestyle=\"\",\n", ")\n", "\n", "plt.show(fig)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The above plot has the standard Matplotlib formatting, with the grid being explicitly added (it is not be default) using the `ax.grid()` method.\n", "\n", "We could have alternatively make the plot with Numpy arrays instead of the data frame as a data source. In that case, our call to `ax.plot()` is\n", "\n", " ax.plot(\n", " df_gfmt.loc[:, \"confidence when correct\"].values,\n", " df_gfmt.loc[:, \"confidence when incorrect\"].values,\n", " marker=\".\",\n", " linestyle=\"\",\n", " )\n", " \n", "If we want to make a plot with various markers and a legend, we build the plot glyph-by-glyph, as with Bokeh. The `label` kwarg of `ax.plot()` specifies the corresponding text in a legend." ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "tags": [] }, "outputs": [ { "data": { "image/svg+xml": [ "\n", "\n", "\n", " \n", " \n", " \n", " \n", " 2023-06-29T14:39:04.331609\n", " image/svg+xml\n", " \n", " \n", " Matplotlib v3.7.1, https://matplotlib.org/\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "\n" ], "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "fig, ax = plt.subplots(figsize=(4, 4))\n", "ax.set_xlabel(\"confidence when correct\")\n", "ax.set_ylabel(\"confidence when incorrect\")\n", "ax.set_xlim(-2.5, 102.5)\n", "ax.set_ylim(-2.5, 102.5)\n", "ax.grid(True)\n", "\n", "ax.plot(\n", " \"confidence when correct\",\n", " \"confidence when incorrect\",\n", " data=df_gfmt.loc[~df_gfmt[\"insomnia\"], :],\n", " marker=\".\",\n", " linestyle=\"\",\n", " label=\"normal sleepers\",\n", ")\n", "\n", "ax.plot(\n", " \"confidence when correct\",\n", " \"confidence when incorrect\",\n", " data=df_gfmt.loc[df_gfmt[\"insomnia\"], :],\n", " marker=\".\",\n", " linestyle=\"\",\n", " color=\"orange\",\n", " label=\"insomniacs\",\n", ")\n", "\n", "ax.legend()\n", "\n", "plt.show(fig)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Aside: Making a scatter plot with Seaborn\n", "\n", "We will take a look at using Seaborn for plotting later in this lesson for making the kinds of plots we have made thus far with iqplot. For now, we demonstrate how to make a scatter plot as above using Seaborn's nifty new (as of June 2023) [objects interface](https://seaborn.pydata.org/tutorial/objects_interface.html). The grammar is similar to Vega-Altair. We have imported `seaborn.objects` as `so`." ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "tags": [] }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "" ] }, "execution_count": 5, "metadata": { "image/png": { "height": 313.22499999999997, "width": 402.9 } }, "output_type": "execute_result" } ], "source": [ "so.Plot(\n", " df_gfmt, \n", " x='confidence when correct', \n", " y='confidence when incorrect',\n", " color='insomnia',\n", ").add(\n", " so.Dot(pointsize=4),\n", ").limit(\n", " x=(-2.5, 102.5),\n", " y=(-2.5, 102.5),\n", ").layout(\n", " size=(4, 4)\n", ").theme(\n", " sns.axes_style(\"whitegrid\"),\n", ")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## A Matplotlib line plot\n", "\n", "We can similarly make a line plot, as we did in [lesson 22](l22_plotting_time_series_generated_data.ipynb). " ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "tags": [] }, "outputs": [ { "data": { "image/svg+xml": [ "\n", "\n", "\n", " \n", " \n", " \n", " \n", " 2023-06-29T14:39:04.879595\n", " image/svg+xml\n", " \n", " \n", " Matplotlib v3.7.1, https://matplotlib.org/\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "\n" ], "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "df_spikes = pd.read_csv(\"data/retina_spikes.csv\", comment=\"#\")\n", "\n", "fig, ax = plt.subplots(figsize=(7, 2))\n", "ax.set_xlabel(\"time (ms)\")\n", "ax.set_ylabel(\"V (µV)\")\n", "ax.grid(True)\n", "\n", "ax.plot(\n", " \"t (ms)\",\n", " \"V (uV)\",\n", " data=df_spikes,\n", " marker=\"\",\n", " linestyle=\"-\",\n", ")\n", "\n", "plt.show(fig)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Plotting with Seaborn\n", "\n", "[Seaborn](https://seaborn.pydata.org/) is a high-level statistical plotting package built on Matplotlib. It is traditionally imported as `sns`, as we have done here. It is best understood by example. We will start by using it to make the same plot of confidence when incorrect versus confidence when correct as we did above (with minor styling differences).\n", "\n", "The `with` block in the code cell below specifies one of [Seaborn's pre-defined styles](https://seaborn.pydata.org/tutorial/aesthetics.html). Note the convenience of using the `hue` kwarg in the call to `sns.scatterplot()`. It automatically colors the points according to the `'insomnia'` column and creates a legend. (Again, it is important that the data frame is tidy!)" ] }, { "cell_type": "code", "execution_count": 7, "metadata": { "tags": [] }, "outputs": [ { "data": { "image/svg+xml": [ "\n", "\n", "\n", " \n", " \n", " \n", " \n", " 2023-06-29T14:39:05.094579\n", " image/svg+xml\n", " \n", " \n", " Matplotlib v3.7.1, https://matplotlib.org/\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "\n" ], "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "with sns.axes_style(\"whitegrid\"):\n", " fig, ax = plt.subplots(figsize=(4, 4))\n", " \n", " ax = sns.scatterplot(\n", " df_gfmt,\n", " x=\"confidence when correct\",\n", " y=\"confidence when incorrect\",\n", " hue=\"insomnia\",\n", " )\n", " \n", " ax.set_xlim(-2.5, 102.5)\n", " ax.set_ylim(-2.5, 102.5)\n", "\n", "plt.show(ax)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "In the plot above, I explicitly instantiated the figure and axes objects so that I could control the size. Seaborn will otherwise do that automatically.\n", "\n", "For the remainder of our plots with Seaborn, we will use the frog tongue strike data set." ] }, { "cell_type": "code", "execution_count": 8, "metadata": { "tags": [] }, "outputs": [], "source": [ "df = pd.read_csv(\"data/frog_tongue_adhesion.csv\", comment=\"#\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We start with a box plot." ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "tags": [] }, "outputs": [ { "data": { "image/svg+xml": [ "\n", "\n", "\n", " \n", " \n", " \n", " \n", " 2023-06-29T14:39:05.316859\n", " image/svg+xml\n", " \n", " \n", " Matplotlib v3.7.1, https://matplotlib.org/\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "\n" ], "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "with sns.axes_style(\"whitegrid\"):\n", " fig, ax = plt.subplots(figsize=(4, 2))\n", "\n", " sns.boxplot(\n", " df,\n", " x=\"impact force (mN)\",\n", " y=\"ID\",\n", " ax=ax,\n", " )\n", " \n", "plt.show(fig)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "To overlay a strip plot, we populate the same axes object using `sns.stripplot()`. We style the box plot so the glyphs do not clash with the strip plot. The `showfliers` kwarg suppresses plotting outliers in the box plot, since those will appear in the strip plot anyway." ] }, { "cell_type": "code", "execution_count": 10, "metadata": { "tags": [] }, "outputs": [ { "data": { "image/svg+xml": [ "\n", "\n", "\n", " \n", " \n", " \n", " \n", " 2023-06-29T14:39:05.520136\n", " image/svg+xml\n", " \n", " \n", " Matplotlib v3.7.1, https://matplotlib.org/\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "\n" ], "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "with sns.axes_style(\"whitegrid\"):\n", " fig, ax = plt.subplots(figsize=(4, 2))\n", "\n", " sns.boxplot(\n", " df,\n", " x=\"impact force (mN)\",\n", " y=\"ID\",\n", " ax=ax,\n", " color=\"white\",\n", " showfliers=False,\n", " showcaps=False,\n", " )\n", " \n", " sns.stripplot(\n", " df, x=\"impact force (mN)\", y=\"ID\", ax=ax, hue=\"ID\", jitter=True, legend=False\n", " )\n", " \n", "plt.show(fig)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Next, we will make a histogram overlayed with a rug plot, which we need to do explicitly with a call to `sns.rugplot()`." ] }, { "cell_type": "code", "execution_count": 11, "metadata": { "tags": [] }, "outputs": [ { "data": { "image/svg+xml": [ "\n", "\n", "\n", " \n", " \n", " \n", " \n", " 2023-06-29T14:39:05.768075\n", " image/svg+xml\n", " \n", " \n", " Matplotlib v3.7.1, https://matplotlib.org/\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "\n" ], "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "with sns.axes_style(\"whitegrid\"):\n", " fig, ax = plt.subplots(figsize=(5, 3))\n", " \n", " sns.histplot(df, x='impact force (mN)', hue='ID', common_bins=False, ax=ax)\n", " sns.rugplot(df, x='impact force (mN)', hue='ID', ax=ax)\n", "\n", "plt.show(fig)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Finally, we can make an ECDF plot." ] }, { "cell_type": "code", "execution_count": 12, "metadata": { "tags": [] }, "outputs": [ { "data": { "image/svg+xml": [ "\n", "\n", "\n", " \n", " \n", " \n", " \n", " 2023-06-29T14:39:05.991855\n", " image/svg+xml\n", " \n", " \n", " Matplotlib v3.7.1, https://matplotlib.org/\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "\n" ], "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "with sns.axes_style(\"whitegrid\"):\n", " fig, ax = plt.subplots(figsize=(5, 3))\n", " \n", " sns.ecdfplot(df, x='impact force (mN)', hue='ID')\n", " \n", "plt.show(fig)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## There's a lot more!\n", "\n", "This is just scratching the surface of what Matplotlib and Seaborn can do. To learn more, you might want to start with their galleries ([Matplotlib](https://matplotlib.org/stable/gallery/index.html), [Seaborn](https://seaborn.pydata.org/examples/index.html))." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Computing environment" ] }, { "cell_type": "code", "execution_count": 13, "metadata": { "tags": [ "hide-input" ] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Python implementation: CPython\n", "Python version : 3.11.3\n", "IPython version : 8.12.0\n", "\n", "pandas : 1.5.3\n", "matplotlib: 3.7.1\n", "seaborn : 0.12.2\n", "jupyterlab: 3.6.3\n", "\n" ] } ], "source": [ "%load_ext watermark\n", "%watermark -v -p pandas,matplotlib,seaborn,jupyterlab" ] } ], "metadata": { "anaconda-cloud": {}, "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.10.11" } }, "nbformat": 4, "nbformat_minor": 4 }