Plotting using jupyter notebooks and the ggplot module in Python¶
- First, install Anaconda (Python 3.6 distribution with jupyter and other data analysis modules included) following the directions here: https://www.continuum.io/downloads
Open Anaconda Prompt (Windows) or Terminal (Mac) and type:
pip install ggplot
Next, start jupyter by typing:
jupyter notebook
- Generate a new notebook by clicking New, followed by Python under Notebooks. Give it a name by clicking on Untitled.
- Test the code that follows in your new notebook. The code is also available in this notebook for your reference.
- For more details on jupyter notebooks, see http://jupyter-notebook-beginner-guide.readthedocs.io/en/latest/index.html
- For more details on ggplot, see http://ggplot.yhathq.com/how-it-works.html
In [1]:
### Import needed modules
#ggplot is used for plotting and is based on the R package ggplot2
#Pandas is a widely-used Python library for statistics, particularly on tabular data. It installed with Anaconda.
#We will use it to read and write tables in .csv format.
from ggplot import *
import pandas
Download the gapminder CSV file¶
http://hwheeler01.github.io/comp150/ggplot/gapminder.csv
- You might need to right-click the link to save.
- Place the file in the same folder as this notebook (
.ipynbfile).
In [3]:
### read in the gapminder data as a pandas DataFrame
gap = pandas.read_csv('gapminder.csv')
In [4]:
### Use DataFrame.info to find out more about a DataFrame
gap.info()
In [5]:
### Use DataFrame.head to view the first few rows
print(gap.head())
In [6]:
### to see more
print(gap.head(n=20))
In [7]:
### to see the last few rows
print(gap.tail())
In [8]:
### Use DataFrame.describe to get summary statistics about continuous data
print(gap.describe())
In [9]:
### Let's initialize a plot
ggplot(gap, aes(x = 'gdpPercap', y = 'lifeExp'))
Out[9]:
In [10]:
### Add points
ggplot(gap, aes(x = 'gdpPercap', y = 'lifeExp')) + geom_point()
Out[10]:
In [11]:
### Let's log transform the x-axis
ggplot(gap, aes(x = 'gdpPercap', y = 'lifeExp')) + geom_point() + scale_x_log()
Out[11]:
In [12]:
### How about some color?
ggplot(gap, aes(x = 'gdpPercap', y = 'lifeExp', color = 'continent')) + geom_point() + scale_x_log()
Out[12]:
In [13]:
### Plot lifeExp vs. year colored by continent
ggplot(gap, aes(x = 'year', y = 'lifeExp', color = 'continent')) + geom_point()
Out[13]:
In [14]:
### Let's separate by continent
ggplot(gap, aes(x = 'year', y = 'lifeExp', color = 'continent')) + geom_point() + facet_wrap('continent')
Out[14]:
In [15]:
### Let's remove Oceania and connect countries with lines
no_oceania = gap[gap.continent != 'Oceania']
ggplot(no_oceania, aes(x = 'year', y = 'lifeExp', color = 'continent', group = 'country')) + \
geom_line() + facet_wrap('continent')
Out[15]:
In [16]:
### Let's pull out a subset of countries to plot
subgap = gap[gap.country.str.contains("Cambodia|Rwanda|United|Japan|Mexico")]
ggplot(subgap, aes(x = 'year', y = 'lifeExp', color = 'country')) + geom_line() + geom_point()
Out[16]:
In [17]:
### We can also do boxplots
ggplot(gap, aes(x = 'continent', y = 'lifeExp')) + geom_boxplot()
Out[17]:
In [18]:
### and change axis labels
ggplot(gap, aes(x = 'continent', y = 'lifeExp')) + geom_boxplot() + xlab("Continent") + ylab("Life Expectancy (years)")
Out[18]:
In [19]:
### and histograms
ggplot(gap, aes(x = 'lifeExp')) + geom_histogram()
Out[19]:
In [20]:
### we can adjust binwidth
for i in range(1,4):
print("binwidth = " + str(i))
print(ggplot(gap, aes(x = 'lifeExp')) + geom_histogram(binwidth = i))
print("\n")
In [21]:
### we can make density plots
ggplot(gap, aes(x = 'lifeExp')) + geom_density()
Out[21]:
In [22]:
### and color by continent
ggplot(gap, aes(x = 'lifeExp', color = 'continent')) + geom_density()
Out[22]:
In [23]:
### we can facet anything
ggplot(no_oceania, aes(x = 'lifeExp',color = 'continent')) + geom_histogram(binwidth = 1) + \
facet_wrap('continent')
Out[23]:
In [24]:
### Let's play with colors!
ggplot(no_oceania, aes(x = 'year', y = 'lifeExp', color = 'continent', group = 'country')) + \
geom_line() + facet_wrap('continent') + scale_color_manual(values = ['red','purple','teal','orange'])
Out[24]:
Print a list of options from the color brewer, see http://colorbrewer2.org/ for visual details.¶
In [25]:
import brewer2mpl
brewer2mpl.print_maps()
In [26]:
ggplot(no_oceania, aes(x = 'year', y = 'lifeExp', color = 'continent', group = 'country')) + \
geom_line() + facet_wrap('continent') + scale_color_brewer(type = 'seq', palette = 'Greens')
Out[26]:
In [27]:
ggplot(no_oceania, aes(x = 'year', y = 'lifeExp', color = 'continent', group = 'country')) + \
geom_line() + facet_wrap('continent') + scale_color_brewer(type = 'div', palette = 'PuOr')
Out[27]:
In [28]:
ggplot(no_oceania, aes(x = 'year', y = 'lifeExp', color = 'continent', group = 'country')) + \
geom_line() + facet_wrap('continent') + scale_color_brewer(type = 'qual', palette = 'Pastel1')
Out[28]:
In [29]:
ggplot(no_oceania, aes(x = 'year', y = 'lifeExp', color = 'continent', group = 'country')) + \
geom_line() + facet_wrap('continent') + scale_color_brewer(type = 'qual', palette = 'Dark2')
Out[29]:
In [30]:
### Change background theme
ggplot(no_oceania, aes(x = 'year', y = 'lifeExp', color = 'continent', group = 'country')) + \
geom_line() + facet_wrap('continent') + scale_color_brewer(type = 'qual', palette = 'Dark2') + \
theme_bw()
Out[30]:
In [ ]: