This article was published as part of the Data Science Blogathon.
Introduction
Data visualization in Python is perhaps one of the most used features for data science with Python today. Libraries in Python come with many different features that allow users to create highly custom graphics, elegant and interactive.
In this article, we will cover the use of Matplotlib, Seaborn, as well as an introduction to other alternative packages that can be used in Python visualization.
Inside Matplotlib and Seaborn, we'll cover some of the most widely used plots in the data science world for easy visualization.
Later in the article, we will go over another powerful feature in Python visualizations, la subtrama, and we covered a basic tutorial for creating subplots.
Useful packages for visualizations in python
Matplotlib
Matplotlib is a Python display library for 2D array diagrams. Matplotlib is written in Python and makes use of the NumPy library. Can be used in Python and IPython shells, Jupyter laptops and web application servers. Matplotlib comes with a wide variety of graphs like line, bar, dispersion, histogram, etc. that can help us deepen our understanding of trends, patterns and correlations. It was introduced by John Hunter in 2002.
Seaborn
Seaborn is a dataset-oriented library for performing statistical representations in Python. It is developed on matplotlib and to create different visualizations. It is integrated with pandas data structures. The library does the mapping and aggregation internally to create informative visuals. It is recommended to use a Jupyter interface / IPython and modo matplotlib.
Bokeh
Bokeh is an interactive display library for modern web browsers. It is suitable for streaming or large data assets and can be used to develop interactive charts and dashboards. There is a wide range of intuitive graphics in the library that can be leveraged to develop solutions. Works closely with PyData tools. The library is suitable for creating custom images according to the required use cases. Images can also be made interactive to serve as a hypothetical scenario model. All code is open source and available on GitHub.
Altair
Altair is a declarative statistical display library for Python. Altair API is easy to use and consistent, and is built on the Vega-Lite JSON specification. The declarative library indicates that when creating any visual object, we need to define the links between the data columns and the channels (X axis, Axis y, size, color). With the help of Altair, informative images can be created with minimal code. Altair has a declarative grammar of both visualization and interaction.
tramadamente
plotly.py is an interactive display library, open source, high level, declarative and browser-based for Python. Contains a variety of useful visualization including scientific charts, 3D graphics, statistical graphs, financial charts, among others. Plot graphics can be viewed in Jupyter notebooks, standalone HTML files or hosted online. Plotly library offers options for interaction and editing. The robust API works perfectly in both web and local browser mode.
ggplot
ggplot is a Python implementation of the graphing grammar. Graphics grammar refers to mapping data to aesthetic attributes (color, shape, size) and geometric objects (points, lines, bars). The basic building blocks according to the grammar of graphs are data, geom (geometric objects), statistics (statistical transformations), scale, coordinate system and facet.
Using ggplot in Python allows you to develop informative visualizations incrementally, understanding the nuances of the data first and then adjusting the components to improve visual representations.
How to use the correct visualization?
To extract the required information from the different visual elements that we create, it is essential that we use the correct representation based on the type of data and the questions we are trying to understand. Then, we will look at a set of most used representations and how we can use them most effectively.
Bar graphic
A bar chart is used when we want to compare metric values in different subgroups of data. If we have a greater number of groups, a bar chart is preferred over a column chart.
Bar chart using Matplotlib
#Creating the datasetdf = sns.load_dataset('titanic') df=df.groupby('who')['fare'].sum().to_frame().reset_index()#Creating the bar chartplt.barh(df['who'],df['fare'],color = ['# F0F8FF','#E6E6FA','#B0E0E6']) #Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') #Show the plotplt.show()
Bar chart with Seaborn
#Creating bar plotsns.barplot(x = 'fare',y = 'who',data = titanic_dataset,palette = "Blues")#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') # Show the plotplt.show()
Column chart
Column charts are mainly used when we need to compare a single category of data between individual sub-items, for instance, when comparing income between regions.
Column chart using Matplotlib
#Creating the datasetdf = sns.load_dataset('titanic') df=df.groupby('who')['fare'].sum().to_frame().reset_index()#Creating the column plot plt.bar(df['who'],df['fare'],color = ['# F0F8FF','#E6E6FA','#B0E0E6']) #Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') #Show the plotplt.show()
Column chart with Seaborn
#Reading the datasettitanic_dataset = sns.load_dataset('titanic')#Creating column chartsns.barplot(x = 'who',y = 'do',data = titanic_dataset,palette = "Blues")#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') # Show the plotplt.show()
Clustered bar chart
A clustered bar chart is used when we want to compare the values in certain groups and subgroups.
Clustered bar chart using Matplotlib
#Creating the datasetdf = sns.load_dataset('titanic')df_pivot = pd.pivot_table(df, values="fare",index="who",columns="class", aggfunc=np.mean)#Creating a grouped bar chartax = df_pivot.plot(kind="bar",alpha=0.5)#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') # Show the plotplt.show()
Bar chart grouped with Seaborn
#Reading the datasettitanic_dataset = sns.load_dataset('titanic')#Creating the bar plot grouped across classessns.barplot(x = 'who',y = 'do',hue="class",data = titanic_dataset, palette = "Blues")#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') # Show the plotplt.show()
Stacked bar chart
A stacked bar chart is used when we want to compare the total sizes of the available groups and the composition of the different subgroups.
Stacked bar chart using Matplotlib
# Stacked bar chart #Creating the datasetdf = pd.DataFrame(columns=["A","B", "C","D"], data=[["E",0,1,1], ["F",1,1,0], ["G",0,1,0]])df.plot.bar(x='A', y =["B", "C","D"], stacked=True, width = 0.4,alpha=0.5) #Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') #Show the plotplt.show()
Stacked Bar Chart with Seaborn
dataframe = pd.DataFrame(columns=["A","B", "C","D"], data=[["E",0,1,1], ["F",1,1,0], ["G",0,1,0]])dataframe.set_index('A').T.plot(kind='bar', stacked=True)#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') # Show the plotplt.show()
Line graph
A line chart is used to represent continuous data points. This visual element can be used effectively when we want to understand the trend over time..
Line chart using Matplotlib
#Creating the datasetdf = sns.load_dataset("iris") df=df.groupby('sepal_length')['sepal_width'].sum().to_frame().reset_index()#Creating the line chartplt.plot(df['sepal_length'], df['sepal_width']) #Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') #Show the plotplt.show()
Line chart with Seaborn
#Creating the datasetcars = ['AUDI', 'BMW', 'NISSAN', 'TESLA', 'HYUNDAI', 'HONDA'] data = [20, 15, 15, 14, 16, 20] #Creating the pie chartplt.pie(data, labels = cars,colors = ['# F0F8FF','#E6E6FA','#B0E0E6','#7B68EE','#483D8B'])#Adding the aestheticsplt.title('Chart title')#Show the plotplt.show()Pie chart
Pie charts can be used to identify proportions of the different components in a given whole..
Pie chart with Matplotlib
#Creating the datasetcars = ['AUDI', 'BMW', 'NISSAN', 'TESLA', 'HYUNDAI', 'HONDA'] data = [20, 15, 15, 14, 16, 20] #Creating the pie chartplt.pie(data, labels = cars,colors = ['# F0F8FF','#E6E6FA','#B0E0E6','#7B68EE','#483D8B'])#Adding the aestheticsplt.title('Chart title')#Show the plotplt.show()
Area chart
Area charts are used to track changes over time for one or more groups. Area charts are preferred over line charts when we want to capture changes over time for more than one group.
Area graph using Matplotlib
#Reading the datasetx=range(1,6)y =[ [1,4,6,8,9], [2,2,7,10,12], [2,8,5,10,6] ]#Creating the area chart ax = plt.gca()ax.stackplot(x, Y, labels=['A','B','C'],alpha=0.5)#Adding the aestheticsplt.legend(loc ="upper left")plt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') #Show the plotplt.show()
Area chart using Seaborn
# Datayears_of_experience =[1,2,3]salary=[ [6,8,10], [4,5,9], [3,5,7] ]# Plotplt.stackplot(years_of_experience,salary, labels=['Company A','Company B','Company C'])plt.legend(loc ="upper left")#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') # Show the plotplt.show()
Column histogram
Column histograms are used to observe the distribution of a single variable with few data points.
Column chart using Matplotlib
#Creating the datasetpenguins = sns.load_dataset("penguins")#Creating the column histogramax = plt.gca()ax.hist(penguins['flipper_length_mm'], color="blue",alpha=0.5, bins=10)#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') #Show the plotplt.show()
Column chart with Seaborn
#Reading the datasetpenguins_dataframe = sns.load_dataset("penguins")#Plotting bar histogramsns.distplot(penguins_dataframe['flipper_length_mm'], kde = False, color="blue", bins=10)#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') # Show the plotplt.show()
Line histogram
Line histograms are used to observe the distribution of a single variable with many data points.
Line Histogram Plot Using Matplotlib
#Creating the datasetdf_1 = np.random.normal(0, 1, (1000, ))density = stats.gaussian_kde(df_1)#Creating the line histogramn, x, _ = plt.hist(df_1, bins=np.linspace(-3, 3, 50), histtype=u'step', density=True) plt.plot(x, density(x))#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') #Show the plotplt.show()
Line Histogram Chart with Seaborn
#Reading the datasetpenguins_dataframe = sns.load_dataset("penguins")#Plotting line histogramsns.distplot(penguins_dataframe['flipper_length_mm'], hist = False, where = True, label="Africa")#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') # Show the plotplt.show()
Scatter plot
Scatter charts can be used to identify relationships between two variables. Can be used effectively in circ*mstances where the dependent variable can have multiple values for the independent variable.
Scatter plot using Matplotlib
#Creating the datasetdf = sns.load_dataset("tips")#Creating the scatter plotplt.scatter(df['total_bill'],df['tip'],alpha=0.5 )#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') #Show the plotplt.show()
Scatter plot using Seaborn
#Reading the datasetbill_dataframe = sns.load_dataset("tips")#Creating scatter plotsns.scatterplot(data=bill_dataframe, x="total_bill", y ="tip")#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') # Show the plotplt.show()
Bubble chart
Scatter charts can be used to represent and show relationships between three variables.
Bubble chart with Matplotlib
#Creating the datasetnp.random.seed(42)N = 100x = np.random.normal(170, 20, N)y = x + np.random.normal(5, 25, N)colors = np.random.rand(N)area = (25 * np.random.rand(N))**2df = pd.DataFrame({ 'X': x, 'AND': Y, 'Colors': colors, "bubble_size":area})#Creating the bubble chartplt.scatter('X', 'AND', s="bubble_size",alpha=0.5, data=df)#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') #Show the plotplt.show()
Bubble chart with Seaborn
#Reading the datasetbill_dataframe = sns.load_dataset("tips")#Creating bubble plotsns.scatterplot(data=bill_dataframe, x="total_bill", y ="tip", hue="size", size="size")#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') # Show the plotplt.show()
Box plot
A box plot is used to show the shape of the distribution, its central value and its variability.
Box plot using Matplotlib
from past.builtins import xrange#Creating the datasetdf_1 = [[1,2,5], [5,7,2,2,5], [7,2,5]]df_2 = [[6,4,2], [1,2,5,3,2], [2,3,5,1]]#Creating the box plotticks = ['A', 'B', 'C']plt.figure()bpl = plt.boxplot(df_1, positions=np.array(xrange(len(df_1)))*2.0-0.4, sym = '', widths=0.6)bpr = plt.boxplot(df_2, positions=np.array(xrange(len(df_2)))*2.0+0.4, sym = '', widths=0.6)plt.plot([], c="#D7191C", label="Label 1")plt.plot([], c="#2C7BB6", label="Label 2")#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') plt.legend()plt.xticks(xrange(0, len(ticks) * 2, 2), ticks)plt.xlim(-2, len(ticks)*2)plt.ylim(0, 8)plt.tight_layout()#Show the plotplt.show()
Box plot using Seaborn
#Reading the datasetbill_dataframe = sns.load_dataset("tips")#Creating boxplotsax = sns.boxplot(x="day", y ="total_bill", hue="smoker", data=bill_dataframe, palette="Set3")#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') # Show the plotplt.show()
Waterfall chart
A waterfall chart can be used to explain the gradual transition in the value of a variable that is subject to increases or decreases.
#Reading the datasettest = pd.Series(-1 + 2 * np.random.rand(10), index=list('abcdefghij'))#Function for makig a waterfall chartdef waterfall(series): df = pd.DataFrame({'pos':np.maximum(series,0),'neg':np.minimum(series,0)}) blank = series.c*msum().shift(1).fillna(0) df.plot(kind='bar', stacked=True, bottom=blank, color=['r','b'], alpha=0.5) step = blank.reset_index(drop=True).repeat(3).shift(-1) step[1::3] = np.nan plt.plot(step.index, step.values,'k')#Creating the waterfall chartwaterfall(test)#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title')#Show the plotplt.show()
diagram of Venn
Venn diagrams are used to see the relationships between two or three sets of elements. Highlight the similarities and differences
from matplotlib_venn import venn3#Making friend chartfriend3(subsets = (10, 8, 22, 6,9,4,2))plt.show()
Tree map
Treemaps are mainly used to display data grouped and nested in a hierarchical structure and to observe the contribution of each component.
import squarify sizes = [40, 30, 5, 25, 10]squarify.plot(sizes)#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') # Show the plotplt.show()
Bar graphic 100% stacked
You can take advantage of a stacked bar chart by 100% when we want to show the relative differences within each group for the different subgroups available.
#Reading the datasetr = [0,1,2,3,4]raw_data = {'greenBars': [20, 1.5, 7, 10, 5], 'orangeBars': [5, 15, 5, 10, 15],'blueBars': [2, 15, 18, 5, 10]}df = pd.DataFrame(raw_data)# From raw value to percentagetotals = [i+j+k for i,j,k in zip(df['greenBars'], df['orangeBars'], df['blueBars'])]greenBars = [i / j * 100 for i,j in zip(df['greenBars'], totals)]orangeBars = [i / j * 100 for i,j in zip(df['orangeBars'], totals)]blueBars = [i / j * 100 for i,j in zip(df['blueBars'], totals)]# plotbarWidth = 0.85names = ('A','B','C','D','E')# Create green Barsplt.bar(r, greenBars, color="#b5ffb9", edgecolor="white", width=barWidth)# Create orange Barsplt.bar(r, orangeBars, bottom=greenBars, color="#f9bc86", edgecolor="white", width=barWidth)# Create blue Barsplt.bar(r, blueBars, bottom=[i+j for i,j in zip(greenBars, orangeBars)], color="#a3acff", edgecolor="white", width=barWidth)# Custom x axisplt.xticks(r, names)plt.xlabel("group")#Adding the aestheticsplt.title('Chart title')plt.xlabel('X axis title')plt.ylabel('Y axis title') plt.show()
Marginal plots
Marginal plots are used to evaluate the relationship between two variables and examine their distributions. Such scatter plots that have histograms, box plots or dot plots on the margins of the respective x and y axes
#Reading the datasetiris_dataframe = sns.load_dataset('iris')#Creating marginal graphssns.jointplot(x=iris_dataframe["sepal_length"], y = iris_dataframe["sepal_width"], kind='scatter')# Show the plotplt.show()
Subparcelas
Subframes are powerful displays that facilitate comparisons between frames
#Creating the datasetdf = sns.load_dataset("iris") df=df.groupby('sepal_length')['sepal_width'].sum().to_frame().reset_index()#Creating the subplotfig, axes = plt.subplots(nrows = 2, ncols = 2)ax=df.plot('sepal_length', 'sepal_width',ax=axes[0,0])ax.get_legend().remove()#Adding the aestheticsax.set_title('Chart title')ax.set_xlabel('X axis title')ax.set_ylabel('Y axis title')ax=df.plot('sepal_length', 'sepal_width',ax=axes[0,1])ax.get_legend().remove()ax=df.plot('sepal_length', 'sepal_width',ax=axes[1,0])ax.get_legend().remove()ax=df.plot('sepal_length', 'sepal_width',ax=axes[1,1])ax.get_legend().remove()#Show the plotplt.show()
In conclusion, there are a variety of different libraries that can be leveraged to their full potential by understanding the use case and requirement. Syntax and semantics vary from package to package and understanding the challenges and benefits of different libraries is essential. Happy viewing!
Aishwarya A
Data scientist and analytics enthusiast
The media shown in this article is not the property of Analytics Vidhya and is used at the author's discretion.
