Highlights<\/h2>\n\n- Using the
yfinance<\/code> libray to get some historical pricing data for stock pricing<\/li>\n- Creating a basic scatterplot using the
pyplot<\/code> module within matplolib<\/li>\n- Considering some of the basic APIs of
matplot<\/code> library<\/li>\n- Using the
matplotlab<\/code> API to customize various parts of our visualization<\/li>\n- Cover some additional APIs in
matplotlib<\/code> for data visualization<\/li>\n<\/ul>\nMatplotlib, PyPlot, and Pandas 101<\/h2>\n
Pandas is a powerful data science toolkit available for Python and is widely used by many modern data-intensive workflows. It is highly optimized to load, process, and output large amounts of data. Pandas come with an incredibly vast number of methods and classes to perform actions on data\u2014including powerful visualization tools built on the matplotlib<\/code> library.<\/p>\nThe matplotlib.pyplot<\/code> module is highly utilized<\/code> when creating visualizations like scatter plots, histograms, and simple line charts in Pandas. Pandas offer some unique methods and features but, ultimately, allows the use of the matplotlib<\/code> APIs as outlined in the official documentation.<\/p>\nWhile creating charts and graphs with matplotlib<\/code> and Pandas is simple\u2014there is a learning curve. The APIs are broad, often abbreviated oddly, and without ample documentation. Fortunately, getting a basic scatter plot to output is simple enough.<\/p>\nGetting Some Data<\/h2>\n
Before we start splashing dots onto the screen we need data. For this article we need data loaded into a Pandas DataFrame class to allow access to the matplotlib.pyplot.plot.scatter<\/code> API. Just as is the case with most things Pandas-related\u2014this is a breeze! In the code below we’ll load some historic stock pricing data and plot it on a scatterplot.<\/p>\nimport yfinance as yf\r\n\r\n# Download historic pricing data\r\nnvda = yf.download('NVDA', period='3mo')\r\n\r\n# Check the result\r\nprint(type(nvda))\r\nprint(nvda)\r\n\r\n# Note the data is loaded as a DataFrame\r\n<class 'pandas.core.frame.DataFrame'>\r\n\r\n# Note the summary of the data\r\n\r\n Open High ... Adj Close Volume\r\nDate ... \r\n2021-04-22 153.750000 154.384995 ... 148.468475 27778800\r\n2021-04-23 149.342499 153.350006 ... 152.617523 22750000\r\n2021-04-26 151.850006 154.789993 ... 154.744537 19779600\r\n2021-04-27 155.785004 156.705002 ... 153.782257 16457200\r\n2021-04-28 153.645004 155.322495 ... 152.732498 20941600\r\n... ... ... ... ... ...\r\n2021-07-16 190.304993 191.570007 ... 181.610001 68822400\r\n2021-07-19 179.152496 190.419998 ... 187.797501 74906000\r\n2021-07-20 187.300003 188.380005 ... 186.119995 43424500\r\n2021-07-21 188.820007 195.270004 ... 194.100006 37044200\r\n2021-07-22 196.419998 198.869995 ... 194.565002 25553593\r\n\r\n[64 rows x 6 columns]\r\n\r\n# Drop all but two data columns\r\nnvda = nvda[['Adj Close', 'Volume']]\r\n\r\n# View result\r\nprint(nvda)\r\n\r\n Adj Close Volume\r\nDate \r\n2021-04-22 148.468475 27778800\r\n2021-04-23 152.617523 22750000\r\n2021-04-26 154.744537 19779600\r\n2021-04-27 153.782257 16457200\r\n2021-04-28 152.732498 20941600\r\n... ... ...\r\n2021-07-16 181.610001 68822400\r\n2021-07-19 187.797501 74906000\r\n2021-07-20 186.119995 43424500\r\n2021-07-21 194.100006 37044200\r\n2021-07-22 194.664993 25614652\r\n\r\n[64 rows x 2 columns]<\/pre>\nNow we have a data frame object consisting of a DateTime index of two columns: Adj Close<\/code> and Volume<\/code>. Now we can plot our data using the built-in pyplot API accessors in Pandas.<\/p>\nPlotting the Data<\/h2>\n
Using the historic pricing data for the Nvidia Corporation ($NVDA) we are going to use the adjusted closing price as our dependent variable (x) and the volume as our independent variable (y). We will specify these as arguments to the DataFrame.plot.scatter method and view our result. Consider the following example code:<\/p>\n
# Create the plot\r\nplot = nvda.plot.scatter(x='Adj Close', y='Volume')\r\n\r\n# show the plot\r\nplot.figure.show()<\/pre>\nThis code does two primary things:<\/p>\n
\n- Tells Python to create a
matplotlib.pyplot.plt<\/code> class object via the Pandas.plot.scatter<\/code> API<\/li>\n- Display the results of the scatter plot as a
.jpg<\/code> image (by default) via the system’s default image viewer.<\/li>\n<\/ol>\nThis results in the following image:<\/p>\n![\"pandas](\"https:\/\/www.alpharithms.com\/wp-content\/uploads\/1257\/pandas-scatterplot-example-one-default-1024x819.jpg\" "\\"pandas-scatterplot-example-one-default\\"")
<\/a>Scatter plot made via the Pandas.plot.scatter API wrapper to the matplotlib.pyplot.plt.scatter class. (Click to enlarge)<\/figcaption><\/figure>\nCustomizing the Scatterplot<\/h2>\n
This scatterplot is enough for quick analysis and can be used to find outliers, non-normal distributions, and identify issues such as multi-collinearity before building linear models<\/a>. This may not be enough for those seeking more aesthetic visualizations. Let’s take a look at some easy ways to customize the scatterplot.<\/p>\nThe scatterplot produced via Pandas makes available all the matplotlib<\/code> scatter plot API options outlined in the official matplotlib documentation<\/a>. Below are some cherry-picked implementations:<\/p>\n# Create the scatter plot with custom arguments\r\nplot = nvda.plot.scatter(\r\n x='Adj Close', y='Volume',\r\n\r\n # <pyplot> class arguments\r\n figsize=(8, 6),\r\n\r\n # <scatter> plot class arguments\r\n s=48,\r\n marker='o',\r\n c='#ff9900',\r\n\r\n # <collection> class arguments\r\n edgecolors='black',\r\n)\r\n\r\n# Remove axis labels via plt.axes API\r\nplot.axes.xaxis.label.set_fontsize(20)\r\nplot.axes.yaxis.label.set_fontsize(20)\r\n\r\n# Custom font + size via plot.axes API\r\nplot.axes.xaxis.label.set_fontfamily('Monospace')\r\nplot.axes.yaxis.label.set_fontfamily('Monospace')\r\nplot.axes.set_xticklabels(labels=[])\r\nplot.axes.set_yticklabels(labels=[])\r\n\r\n# Show the plot\r\nplot.figure.show()<\/pre>\nThis code does several things in a cumbersome way:<\/p>\n
\n- Customizes the size of the image to 8″ x 6″<\/li>\n
- Increases the marker size<\/li>\n
- Specifies a custom marker type<\/li>\n
- Creates a custom color for the marker<\/li>\n
- Outlines the marker in black<\/li>\n
- Removes the values from the X and Y axes<\/li>\n
- Customizes font + size of the axis labels<\/li>\n<\/ol>\n
These specifications result in the following scatterplot:<\/p>\n![\"pandas](\"https:\/\/www.alpharithms.com\/wp-content\/uploads\/1257\/pandas-scatterplot-customization-1024x819.jpg\" "\\"pandas-scatterplot-customization\\"")
<\/a>A Pandas-generated scatterplot with customized parameters for font, axes labels, marker color, shape, and size. (click to enlarge)<\/figcaption><\/figure>\nFor more ideas on how to customize the Pandas scatter plot (and other plot types) checkout the methods outlined in the following matplotlib<\/code> documentation files:<\/p>\n\n- PyPlot API<\/a><\/li>\n
- Scatter API<\/a><\/li>\n
- Collections API<\/a><\/li>\n<\/ul>\n
Review<\/h2>\n
Generating scatter plots in Python is simple using the Pandas<\/code> or matplotlib<\/code> library. Given the Pandas library utilizes the matplotlib<\/code> library under the hood\u2014it pays to be familiar with its APIs if seeking to create customized visualizations. For ideas on other cool ways to visualize data with Pandas check out our article about autocorrelation and time series plots<\/a> as well as our article about creating heatmaps for correlation matrices in Python<\/a>. With access to so many powerful data analysis and visualization tools it’s no wonder that Python continues to reign atop the lists of most popular programming languages<\/a> year after year.<\/p>\n","protected":false},"excerpt":{"rendered":"Scatterplots are incredibly useful visualization tools for visualizing and analyzing data. The Pandas library in Python makes creating, saving, and customizing scatter plots incredibly simple. Whether using scatter plots for initial inspection or more involved processes like linear regression\u2014Pandas and Python make things a breeze! In this article, we’ll walk through the process of getting […]<\/p>\n","protected":false},"author":2,"featured_media":1260,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"inline_featured_image":false,"footnotes":""},"categories":[655,10],"tags":[398,192,191,459,210,383,143,193,212],"acf":[],"_links":{"self":[{"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/posts\/1257"}],"collection":[{"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/comments?post=1257"}],"version-history":[{"count":4,"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/posts\/1257\/revisions"}],"predecessor-version":[{"id":1555,"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/posts\/1257\/revisions\/1555"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/media\/1260"}],"wp:attachment":[{"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/media?parent=1257"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/categories?post=1257"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/tags?post=1257"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
yfinance<\/code> libray to get some historical pricing data for stock pricing<\/li>\n- Creating a basic scatterplot using the
pyplot<\/code> module within matplolib<\/li>\n- Considering some of the basic APIs of
matplot<\/code> library<\/li>\n- Using the
matplotlab<\/code> API to customize various parts of our visualization<\/li>\n- Cover some additional APIs in
matplotlib<\/code> for data visualization<\/li>\n<\/ul>\nMatplotlib, PyPlot, and Pandas 101<\/h2>\n
Pandas is a powerful data science toolkit available for Python and is widely used by many modern data-intensive workflows. It is highly optimized to load, process, and output large amounts of data. Pandas come with an incredibly vast number of methods and classes to perform actions on data\u2014including powerful visualization tools built on the matplotlib<\/code> library.<\/p>\nThe matplotlib.pyplot<\/code> module is highly utilized<\/code> when creating visualizations like scatter plots, histograms, and simple line charts in Pandas. Pandas offer some unique methods and features but, ultimately, allows the use of the matplotlib<\/code> APIs as outlined in the official documentation.<\/p>\nWhile creating charts and graphs with matplotlib<\/code> and Pandas is simple\u2014there is a learning curve. The APIs are broad, often abbreviated oddly, and without ample documentation. Fortunately, getting a basic scatter plot to output is simple enough.<\/p>\nGetting Some Data<\/h2>\n
Before we start splashing dots onto the screen we need data. For this article we need data loaded into a Pandas DataFrame class to allow access to the matplotlib.pyplot.plot.scatter<\/code> API. Just as is the case with most things Pandas-related\u2014this is a breeze! In the code below we’ll load some historic stock pricing data and plot it on a scatterplot.<\/p>\nimport yfinance as yf\r\n\r\n# Download historic pricing data\r\nnvda = yf.download('NVDA', period='3mo')\r\n\r\n# Check the result\r\nprint(type(nvda))\r\nprint(nvda)\r\n\r\n# Note the data is loaded as a DataFrame\r\n<class 'pandas.core.frame.DataFrame'>\r\n\r\n# Note the summary of the data\r\n\r\n Open High ... Adj Close Volume\r\nDate ... \r\n2021-04-22 153.750000 154.384995 ... 148.468475 27778800\r\n2021-04-23 149.342499 153.350006 ... 152.617523 22750000\r\n2021-04-26 151.850006 154.789993 ... 154.744537 19779600\r\n2021-04-27 155.785004 156.705002 ... 153.782257 16457200\r\n2021-04-28 153.645004 155.322495 ... 152.732498 20941600\r\n... ... ... ... ... ...\r\n2021-07-16 190.304993 191.570007 ... 181.610001 68822400\r\n2021-07-19 179.152496 190.419998 ... 187.797501 74906000\r\n2021-07-20 187.300003 188.380005 ... 186.119995 43424500\r\n2021-07-21 188.820007 195.270004 ... 194.100006 37044200\r\n2021-07-22 196.419998 198.869995 ... 194.565002 25553593\r\n\r\n[64 rows x 6 columns]\r\n\r\n# Drop all but two data columns\r\nnvda = nvda[['Adj Close', 'Volume']]\r\n\r\n# View result\r\nprint(nvda)\r\n\r\n Adj Close Volume\r\nDate \r\n2021-04-22 148.468475 27778800\r\n2021-04-23 152.617523 22750000\r\n2021-04-26 154.744537 19779600\r\n2021-04-27 153.782257 16457200\r\n2021-04-28 152.732498 20941600\r\n... ... ...\r\n2021-07-16 181.610001 68822400\r\n2021-07-19 187.797501 74906000\r\n2021-07-20 186.119995 43424500\r\n2021-07-21 194.100006 37044200\r\n2021-07-22 194.664993 25614652\r\n\r\n[64 rows x 2 columns]<\/pre>\nNow we have a data frame object consisting of a DateTime index of two columns: Adj Close<\/code> and Volume<\/code>. Now we can plot our data using the built-in pyplot API accessors in Pandas.<\/p>\nPlotting the Data<\/h2>\n
Using the historic pricing data for the Nvidia Corporation ($NVDA) we are going to use the adjusted closing price as our dependent variable (x) and the volume as our independent variable (y). We will specify these as arguments to the DataFrame.plot.scatter method and view our result. Consider the following example code:<\/p>\n
# Create the plot\r\nplot = nvda.plot.scatter(x='Adj Close', y='Volume')\r\n\r\n# show the plot\r\nplot.figure.show()<\/pre>\nThis code does two primary things:<\/p>\n
\n- Tells Python to create a
matplotlib.pyplot.plt<\/code> class object via the Pandas.plot.scatter<\/code> API<\/li>\n- Display the results of the scatter plot as a
.jpg<\/code> image (by default) via the system’s default image viewer.<\/li>\n<\/ol>\nThis results in the following image:<\/p>\n<\/a>Scatter plot made via the Pandas.plot.scatter API wrapper to the matplotlib.pyplot.plt.scatter class. (Click to enlarge)<\/figcaption><\/figure>\nCustomizing the Scatterplot<\/h2>\n
This scatterplot is enough for quick analysis and can be used to find outliers, non-normal distributions, and identify issues such as multi-collinearity before building linear models<\/a>. This may not be enough for those seeking more aesthetic visualizations. Let’s take a look at some easy ways to customize the scatterplot.<\/p>\nThe scatterplot produced via Pandas makes available all the matplotlib<\/code> scatter plot API options outlined in the official matplotlib documentation<\/a>. Below are some cherry-picked implementations:<\/p>\n# Create the scatter plot with custom arguments\r\nplot = nvda.plot.scatter(\r\n x='Adj Close', y='Volume',\r\n\r\n # <pyplot> class arguments\r\n figsize=(8, 6),\r\n\r\n # <scatter> plot class arguments\r\n s=48,\r\n marker='o',\r\n c='#ff9900',\r\n\r\n # <collection> class arguments\r\n edgecolors='black',\r\n)\r\n\r\n# Remove axis labels via plt.axes API\r\nplot.axes.xaxis.label.set_fontsize(20)\r\nplot.axes.yaxis.label.set_fontsize(20)\r\n\r\n# Custom font + size via plot.axes API\r\nplot.axes.xaxis.label.set_fontfamily('Monospace')\r\nplot.axes.yaxis.label.set_fontfamily('Monospace')\r\nplot.axes.set_xticklabels(labels=[])\r\nplot.axes.set_yticklabels(labels=[])\r\n\r\n# Show the plot\r\nplot.figure.show()<\/pre>\nThis code does several things in a cumbersome way:<\/p>\n
\n- Customizes the size of the image to 8″ x 6″<\/li>\n
- Increases the marker size<\/li>\n
- Specifies a custom marker type<\/li>\n
- Creates a custom color for the marker<\/li>\n
- Outlines the marker in black<\/li>\n
- Removes the values from the X and Y axes<\/li>\n
- Customizes font + size of the axis labels<\/li>\n<\/ol>\n
These specifications result in the following scatterplot:<\/p>\n<\/a>A Pandas-generated scatterplot with customized parameters for font, axes labels, marker color, shape, and size. (click to enlarge)<\/figcaption><\/figure>\nFor more ideas on how to customize the Pandas scatter plot (and other plot types) checkout the methods outlined in the following matplotlib<\/code> documentation files:<\/p>\n\n- PyPlot API<\/a><\/li>\n
- Scatter API<\/a><\/li>\n
- Collections API<\/a><\/li>\n<\/ul>\n
Review<\/h2>\n
Generating scatter plots in Python is simple using the Pandas<\/code> or matplotlib<\/code> library. Given the Pandas library utilizes the matplotlib<\/code> library under the hood\u2014it pays to be familiar with its APIs if seeking to create customized visualizations. For ideas on other cool ways to visualize data with Pandas check out our article about autocorrelation and time series plots<\/a> as well as our article about creating heatmaps for correlation matrices in Python<\/a>. With access to so many powerful data analysis and visualization tools it’s no wonder that Python continues to reign atop the lists of most popular programming languages<\/a> year after year.<\/p>\n","protected":false},"excerpt":{"rendered":"Scatterplots are incredibly useful visualization tools for visualizing and analyzing data. The Pandas library in Python makes creating, saving, and customizing scatter plots incredibly simple. Whether using scatter plots for initial inspection or more involved processes like linear regression\u2014Pandas and Python make things a breeze! In this article, we’ll walk through the process of getting […]<\/p>\n","protected":false},"author":2,"featured_media":1260,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"inline_featured_image":false,"footnotes":""},"categories":[655,10],"tags":[398,192,191,459,210,383,143,193,212],"acf":[],"_links":{"self":[{"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/posts\/1257"}],"collection":[{"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/comments?post=1257"}],"version-history":[{"count":4,"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/posts\/1257\/revisions"}],"predecessor-version":[{"id":1555,"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/posts\/1257\/revisions\/1555"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/media\/1260"}],"wp:attachment":[{"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/media?parent=1257"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/categories?post=1257"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.alpharithms.com\/wp-json\/wp\/v2\/tags?post=1257"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}