Using a Library for Scientific Plotting

A library is a collection of code units (functions, classes, variables) which implement a set of related tasks. In Python, matplotlib is the most widely used library for scientific plotting. We commonly use its pyplot module and import it as follows:

import matplotlib.pyplot as plt

In Python, a module is an object which contains Python code and which can be imported inside another Python program. Simple plots are then (fairly) simple to create:

time = [0, 1, 2, 3]
position = [0, 100, 200, 300]

plt.plot(time, position)
plt.xlabel('Time (hr)')
plt.ylabel('Position (km)')

Simple Position-Time Plot

The Jupyter notebook renders plots inline by default.

Displaying All Open Figures

In our Jupyter notebook example, running the cell should generate the figure directly below the code. However, other Python environments (such as an interactive Python session started from a terminal or a Python script executed via the command line) require an additional command to display the figure:

plt.show()

This command can also be used within a notebook – for instance, to display multiple figures if several are created by a single cell.

Plotting Data Directly from a Dataframe

In Python, pandas is is the most widely used library for data analysis and manipulation. Notably, it provides a very useful data structure called a dataframe. Dataframes contain tabular data which are:

To showcase the power and ease of use of pandas, let’s try it on some oceanographic open data: https://data.gov.ie/dataset/celtic-explorer-underway Navigate to “available as CSV” and right-click the “Download” button (to copy the URL). We are passing this URL to pandas’s reading function as is, and it just works! Alternatively, we could pass the path to a local data file (stored on our machine). Note (from the file extension) that the data file is a zipped CSV file, so pandas takes care of unzipping it first. CSV stands for “comma-separated values” and it is a very common format for storing tabular data. This open format is both human-readable and machine-readable. Sometimes, instead of commas, values are separated by tabs (TSV) or semicolons.

url = "http://atlas.marine.ie/midata/OceanFeatures/Celtic_Explorer_Underway.csv.zip"

import pandas as pd

df = pd.read_csv(url)
df.shape

(20845, 28)

What are the columns of this table?

df.columns

Index(['WKT', 'gml_id', 'OBJECTID', 'beginLifespanVersion',
       'endLifespanVersion', 'datasetName', 'localId', 'inspireId',
       'inspireTheme', 'version', 'description', 'longitude', 'latitude',
       'depth', 'depthUoM', 'value_', 'observedProperty',
       'observedPropertyUoM', 'resultQuality', 'resultTime', 'validTime',
       'process', 'scale', 'coordinateReferenceSystem',
       'styleLayerDescription', 'licence', 'dataProvenance', 'dataAccessUrl'],
      dtype='object')

We can access a given column using square brackets and the name (label) of this column in quotes. Each column is a series of data:

type(df['value_'])

pandas.core.series.Series

What is this dataset about?

df['description'].unique()

array(['Underway SBE 21 thermosalinograph sea surface temperature station'],
      dtype=object)

This dataset contains values of sea surface temperature (SST):

df['observedProperty'].unique()

array(['http://vocab.nerc.ac.uk/collection/P07/current/CFSN0381/'],
      dtype=object)

We can plot them readily by calling df['value_'].plot(). By default, this function plots the rows as the x-axis.

Using Different Kinds of Plots

Many kinds of plots are available (see the official docs). For example, we may want to know how the SST values are distributed. We would then use a histogram:

# df['value_'].plot(kind='hist')
df['value_'].hist()

SST Histogram

This implicitly uses matplotlib.pyplot.

Styling Plots

Different styles are also available; to view them, run:

plt.style.available

['Solarize_Light2',
 '_classic_test_patch',
 '_mpl-gallery',
 '_mpl-gallery-nogrid',
 'bmh',
 'classic',
 'dark_background',
 'fast',
 'fivethirtyeight',
 'ggplot',
 'grayscale',
 'seaborn-v0_8',
 'seaborn-v0_8-bright',
 'seaborn-v0_8-colorblind',
 'seaborn-v0_8-dark',
 'seaborn-v0_8-dark-palette',
 'seaborn-v0_8-darkgrid',
 'seaborn-v0_8-deep',
 'seaborn-v0_8-muted',
 'seaborn-v0_8-notebook',
 'seaborn-v0_8-paper',
 'seaborn-v0_8-pastel',
 'seaborn-v0_8-poster',
 'seaborn-v0_8-talk',
 'seaborn-v0_8-ticks',
 'seaborn-v0_8-white',
 'seaborn-v0_8-whitegrid',
 'tableau-colorblind10']

plt.style.use('ggplot')
df['value_'].hist();
plt.xlabel('SST (Degrees Celsius)');

SST Histogram with ggplot styling

We have added semicolons to keep only the actual plot in the output.

Alternatively, we could call matplotlib.pyplot’s plot function explicitly:

plt.hist(df['value_']);

SST Histogram from plt.plot

The colour can also be specified as an additional optional argument, e.g., 'b' stands for blue:

df['value_'].hist(color='b');

SST Histogram in blue

Changing Colours

  1. Change the histogram above so it is green instead of blue.
  2. (Re-)add a label for the x-axis (xlabel).
  3. Add a title for the overall plot (title).

Solution

df['value_'].hist(color='g');
plt.xlabel('SST (Degrees Celsius)');
plt.title('SST Measurements from the Celtic Explorer Underway');

SST Histogram in green

Saving your Plot to a File

If you are satisfied with the plot you see, you may want to save it to a file, perhaps to include it in a publication. There is a function in the matplotlib.pyplot module that accomplishes this: savefig. Calling this function, e.g., with

plt.savefig('my_figure.png')

will save the current figure to the file my_figure.png. The file format will be automatically deduced from the filename extension (other formats are pdf, ps, eps, and svg).

Note that functions in plt refer to a global figure variable, and after a figure has been displayed to the screen (e.g., with plt.show), matplotlib will make this variable refer to a new empty figure. Therefore, make sure you call plt.savefig before the plot is displayed to the screen, otherwise you may end up with a file showing an empty plot.

When using dataframes, data is often generated and plotted to screen in one line, so plt.savefig seems not to work readily. One possibility is to:

data.plot(kind='bar')
fig = plt.gcf()  # get current figure
fig.savefig('my_figure.png')

Choosing Data Structure and Plot Type

  1. We can create a dataframe by specifying the content of its columns: table = pd.DataFrame({'Time': time, 'Position': position}). What does table.plot(); produce?
  2. Plot Position against Time. One way of achieving it is table.plot(x='Time', y='Position');. Another way is table.set_index('Time').plot();: Note that this turns the table['Position'] series into a timeseries, i.e., a series indexed in time order. In which cases would you favour one or the other?
  3. What about table.plot(kind='scatter', x='Time', y='Position');?

Solution

  1. Both series are represented against their indices.
  2. Setting the Time column as the dataframe index might be useful if we wanted to include other data: 
    table.set_index('Time', inplace=True)
table['Position_2'] = [300, 400, 200, 100]
table.plot();
  1. A scatter plot is more suitable for representing data points. We could display a regression line to show the relationship between the two series (even if, in this case, it would be trivial!).

Making your Plots Accessible

Whenever you are generating plots to go into a paper or a presentation, there are a few things you can do to make sure that everyone can understand your plots.