Example - Python in Chemistry

In this lesson, we will illustrate the capabilities of Python by running and editing pre-existing code.

This program:

Opens a .csv file containing UV-Vis absorption data
Converts its wavelength units to energy units
Finds all peaks in the data above a threshold
Plots the spectrum and peaks to a scientific standard
Saves the figure as a .pdf

Preparing your folder¶

The easiest way to allow Python programs to interact with files is to store the source code in the same folder as the file. In this case, we need our .py file and .csv files in the same place.

Syntax¶

What now follows is a program which reads all the lines of this file which start with a number (allowing us to skip the header). The data is stored by the program and the last x and y data points are printed out as a health check.

"""
UV spectrum analyser

Input a spectrum in nm units and csv format.

"""

import matplotlib.pyplot as plt

# parse csv data
file_name = "benzene_uv-vis_nm.csv"

wavelengths = []
absorbances = []

with open(file_name) as spec_file:
    for line in spec_file:
        # only conserve lines that start with a number
        # to skip header
        if line[0].isdigit():
            split_line = line.split(",")
            wavelengths.append(float(split_line[0]))
            absorbances.append(float(split_line[1]))

print(wavelengths[-1],absorbances[-1])

Let’s look at a few features of how Python is written, using this as an example.

Top to bottom: Python code is made up of commands for the computer to follow. By default, they are ordered from top to bottom (the first line is executed first, and so on). There are ways to override this rule for our benefit, for example repeating a same line of code hundreds of times without needing to write it repeatedly.
Indentation: The space between the left of the screen and the beginning of Python code is called an indent. In Python, the indentation has a precise meaning for the operation of the code, so it can’t be changed freely to make the program look neater. Generally, indentation is used to modify the top-to-bottom operation of Python.
Colour: Code editors highlight words using different colours to indicate their role in the Python language. It’s as if in English, one colour was reserved for verbs and another for nouns. Some words are invented by the programmer, (like file_name_), but others are chosen from pre-existing words with meaning (like import or open). Syntax highlighting allows you to notice when you are choosing a word which is already reserved in Python.
Comments: Programs can be hard to understand, even by experts. Therefore programmers leave explanatory text to help the reader understand what is going on. These are called comments. Comments can be written by preceding them with a hash symbol (#) or by encasing multiple lines by two sets of triple quotes (""").
Case: Python is case sensitive, which means that capital and lower case letters are not interchangeable. Therefore, while the print() function is reserved and defined in Python, Print() is not.

Debugging code¶

Let’s complete our code.

Task

Under your first bit of code, paste the block below, run your program, and press the “Plots” tab above the output pane in Spyder to see the result.

# covert nm to eV
conversion_factor = 1239.84
energies = []

for wavelength in wavelengths:
    energies.append(conversion_factor/wavelength)

# find peaks
threshold = 3.0
peak_energies = []
peak_absorbances = []

# loop over each point except first and last
for i in range(len(absorbances)-4):
    # find peaks based on adjacent values
    if (absorbances[i+1] < absorbances[i+2]) and (absorbances[i+3] < absorbances[i+2]):
        # apply threshold
        if absorbances[i+2] > threshold:
            peak_absorbances.append(absorbances[i+2])
            peak_energies.append(energies[i+2])

# plot results
fig, ax = plt.subplots()
ax.plot(energies,absorbances,label="Benzene absorption",zorder=1)
ax.scatter(peak_energies,peak_absorbances,color='orange',label="Peaks",zorder=2)
ax.hlines(threshold,energies[0],energies[-1],linestyle='--',color='orange',label="Threshold")
ax.set_xlabel("Energy / eV")
ax.set_ylabel("Absorbance")
ax.legend()
plt.show()

Using the comments, try to understand what each section of the code accomplishes.

You should see a plot on the top right of the window. The spectrum looks fine but our peak detection seems to have failed. Our strategy was to look at each absorbance value and check that it is strictly higher than the value before and after.

Let’s fix our code with a better peak detection strategy.

Task

Replace this line:

    if (absorbances[i+1] < absorbances[i+2]) and (absorbances[i+3] < absorbances[i+2]):

with the following:

    if (absorbances[i] < absorbances[i+2]) and (absorbances[i+4] < absorbances[i+1]):

Taking care to include the leading spaces before if.

Run your code and check that more correct peaks are detected.

Now that our code works as intended, let’s take control of it.

Extra functionality¶

We’ve demonstrated how to use Python in basic data processing. You can already take this code and modify it to suit the uses of your own data.

Let’s demonstrate some last tricks that are possible in Python.

Task

Paste the following line at the bottom of your script and run it:


plt.savefig("spectrum.pdf")

Now search your working folder for a file called spectrum.pdf and open it. Try changing the .pdf extension to other file types to see which ones work.

Finally, let’s use a more complicated piece of code.

Task

Paste the following code at the bottom of your script and run it:

import plotly.graph_objects as go

fig = go.Figure()

# plot line
fig.add_trace(go.Scatter(
    x=energies,
    y=absorbances,
    mode='lines',
    name='Benzene absorption'
))

# plot peaks
fig.add_trace(go.Scatter(
    x=peak_energies,
    y=peak_absorbances,
    mode='markers',
    marker=dict(color='orange'),
    name='Peaks'
))

# plot threshold
fig.add_shape(
    type='line',
    x0=energies[0],
    x1=energies[-1],
    y0=threshold,
    y1=threshold,
    line=dict(color='orange', dash='dash'),
)

# add labels
fig.update_layout(xaxis_title="Energy / eV",
                  yaxis_title="Absorbance",showlegend=True)

fig.show(renderer="browser")

Play around with this interactive visualisation of your data. You can even save it as .html and put it on your own webpage!

Summary¶

Python can interact with files, analyse data, and plot it.
It is executed top-to-bottom.
Indentation is meaningful and can break the rule above.
Syntax highlighting can help you read code.
Comments are written with # or """.
Python is case-sensitive.

The next lessons will return to basics and teach you the fundamentals of Python. By the end, you should be able to understand everything shown in this lesson’s example.

References¶

Romand, J., & Vodar, B. (1951). Spectres d’absorption du benzène à l’état vapeur et à l’état condensé dans l’ultraviolet lointain. Comptes Rendus, 233, 930–932.