Prerequisite: Data Visualization in Python
Visualization is seeing the data along various dimensions. In python, we can visualize the data using various plots available in different modules.
In this article, we are going to visualize and predict the crop production data for different years using various illustrations and python libraries.
Dataset
The Dataset contains different crops and their production from the year 2013 – 2020.
Requirements
There are a lot of python libraries which could be used to build visualization like matplotlib, vispy, bokeh, seaborn, pygal, folium, plotly, cufflinks, and networkx. Of the many, matplotlib and seaborn seems to be very widely used for basic to intermediate level of visualizations.
However, two of the above are widely used for visualization i.e.
- Matplotlib: It is an amazing visualization library in Python for 2D plots of arrays, It is a multi-platform data visualization library built on NumPy arrays and designed to work with the broader SciPy stack. Use the below command to install this library:
pip install matplotlib
- Seaborn: This library sits on top of matplotlib. In a sense, it has some flavors of matplotlib while from the visualization point, it is much better than matplotlib and has added features as well. Use the below command to install this library:
pip install seaborn
Step-by-step Approach
- Import required modules
- Load the dataset.
- Display the data and constraints of the loaded dataset.
- Use different methods to visualize various illustrations from the data.
Visualizations
Below are some programs which indicates the data and illustrates various visualizations of that data:
Example 1:
Python3
# importing pandas moduleimport pandas as pd# load the datasetdata = pd.read_csv('crop.csv')# display top 5 valuesdata.head() |
Output:
These are the top 5 rows of the dataset used.
Example 2:
Python3
# data descriptiondata.info() |
Output:
These are the data constraints of the dataset.
Example 3:
Python3
# 2011 crop data in histogram analysisdata['2011'].hist() |
Output:
The above program depicts the crop production data in the year 2011 using histogram.
Example 4:
Python3
# 2012 crop data in histogram analysisdata['2012'].hist() |
Output:
The above program depicts the crop production data in the year 2012 using histogram.
Example 4:
Python3
# 2013 crop data in histogram analysisdata['2013'].hist() |
Output:
The above program depicts the crop production data in the year 2013 using histogram.
Example 5:
Python3
# display all year datadata.hist() |
Output:
The above program depicts the crop production data of all the available time periods(year) using multiple histograms.
Example 6:
Python3
# import seaborn moduleimport seaborn as sns# setting stylesns.set_style("whitegrid")# plotting data using boxplot for 2013 - 2014sns.boxplot(x='2013', y='2014', data=data) |
Output:
Comparing crop productions in the year 2013 and 2014 using box plot.
Example 7:
Python3
# scatter plot 2013 data vs 2014 dataplt.scatter(data['2013'],data['2014'])plt.show() |
Output:
Comparing crop production in the year 2013 and 2014 using scatter plot.
Example 8:
Python3
# line plot 2013 data vs 2014 dataplt.plot(data['2013'],data['2014'])plt.show() |
Output:
Comparing crop productions in the year 2013 and 2014 using line plot.
Example 9:
Python3
# import required modulesimport matplotlib.pyplot as pltfrom scipy import stats# assign datax = data['2017']y = data['2018']# linear regression 2017 data vs 2018 dataslope, intercept, r, p, std_err = stats.linregress(x, y)# function to return slopedef myfunc(x): return slope * x + interceptmymodel = list(map(myfunc, x))# scatterplt.scatter(x, y)# plotting the dataplt.plot(x, mymodel)# display the figureplt.show() |
Output:
Applying linear regression to visualize and compare predicted crop production data between the year 2017 and 2018.
Example 10:
Python3
# import required modulesimport matplotlib.pyplot as pltfrom scipy import stats# assign datax = data['2016']y = data['2017']# linear regression 2017 data vs 2018 dataslope, intercept, r, p, std_err = stats.linregress(x, y)# function to return slopedef myfunc(x): return slope * x + interceptmymodel = list(map(myfunc, x))# scatterplt.scatter(x, y)# plotting the dataplt.plot(x, mymodel)# display the figureplt.show() |
Output:
Applying linear regression to visualize and compare predicted crop production data between the year 2016 and 2017.
Demo Video
This video shows how to depict the above data visualization and predict data, using Jupyter Notebook from scratch.
In this way various data visualizations and predictions can be computed.
