from IPython.core.display import HTML

def _set_css_style(css_file_path):
   """
   Read the custom CSS file and load it into Jupyter.
   Pass the file path to the CSS file.
   """

   styles = open(css_file_path, "r").read()
   s = '<style>%s</style>' % styles     
   return HTML(s)

_set_css_style('rise.css')

Iteration and data visualization

print view
notebook

• Logical operators and conditional statements
• Group membership
• Loops and iteration
• Slicing
• Data visualization with matplotlib and seaborn

Conditionals and control flow

if statements

• An if statement conditionally executes a block of code
• The condition is some Boolean value
• A block of code is delineated by consistent indentation.
• Blank lines and comments will not end a block
• Whitespace is significant - do not mix tabs and spaces
• In an if..elif..else statement, only one block of code will be executed

if False:
    print(1)
elif True:
    print(2)
else:
    print(3)

What is val?

val = 0
if val >= 0:
    val += 1
elif val < 1:
    val += 2
elif True:
    val += 3
else:
    val += 5
    
    val += 7

val

pass

pass does nothing

if True:
    pass
else:
    print("I will never print")

Logical operators

• and, or, and not
• not has higher precedence than and which has higher precedence than or
• all three have lower precedence than comparison operators
• but use parentheses to avoid confusion

Compound example: not (x and y) = (not x) or (not y)

A more complex example

x = 3; y = 4; z = 5
(x != 3 or not y < 0) and (z < 10 or not y == 0)

print(x != 3)
print(not y < 0)
print(z < 10)
print(not y == 0) 

Membership testing

in and not in can be used to see if an object is in a collection

l = [1, 2, 3]

1 in l

345 in l

"1" not in l

"good" in "goodness"

Membership depends on the container

'cat' in 'catalog'

'cat' in ['catalog', 'catastrophe']

Loops

i = 0
while i < 3:  # as long as this condition is true...
    print(i)  # ...execute this block of code
    i += 1   

letters = ['a','b','c','d','e','f','g']
for i in letters: # for every item in this collection...
    print(i)      # ...execute this block of code with i set to the object

range

The preferred method of looping is to use a for loop. There are a number of builtin functions to help create collections to iterate over.

list(range(3))

list(range(1,10))

list(range(1,10,3))

What is val?

val = 0
for i in range(10):
    val += i

val

Iterable

for will loop over any iterable object

In addition to collections, this includes any object that can provide items one at a time (e.g., files)

f = open("../files/brca1.fasta")
count = 0
for line in f.readlines():
    count += len(line)
count

f = open("../files/brca1.fasta")
count = 0
for line in f:  # can iterate through file object directly!
    count += len(line)
count

break and continue

break will exit out of the entire loop while continue will skip to the next iteration

i = 0
while True:
    i += 1
    if i < 6:
        continue
    print(i)
    if i > 4:
        break

What is the value of x?

x = 0
while x < 100:
    if x < 50:
        pass
    elif x == 5:
        break
    x += 1

x

Slicing

What if you don't want to iterate over the full list?

s = "Hello"
s[0:2]

s[2:] # empty indices indicate remainder of string/list

s[:-2] # negative indices count from the back

s2 = "Hello, world"
s2[:-2]

Slicing a list

l = ["cat", "dog", "fish", "mouse"]
l[2:3]

matplotlib

Makes easy things easy and hard things possible

plt has hundreds of functions for creating and manipulating plots

plt.plot creates line/point graphs

You can specify an arbitrary number of x, y, format arguments

import matplotlib.pyplot as plt
# next is a python notebook command to embed plot output in the notebook
# this may or may not be necessary depending on versions
%matplotlib inline 

plt.plot([0,1,2,3], [0,1,2,3], '-b', [0,1,2,3], [0,1,4,9], '--r')
plt.ylabel('Y')
plt.xlabel('X')
plt.show() # not actually necessary in notebook

matplotlib can overlay multiple plots on the same figure

plt.plot([1,2,4,9,16], label='one') # if only one data set is provide, assume uniform x
plt.plot([0.5,1,2,4.5,8], '^-.', label='two') # keep adding to current graph until show is called
plt.legend() # uses labels
plt.show()

plt.show() # this doesn't show anything since we haven't created another plot
plt.close() # another way to finish with a plot

If you use the %matplotlib command in ipython, the plot will be updated with every command (can skip show)

Many options for formatting plots (marker type, color, size, ...)

Can also specify colors with hexadecimal or color-generating functions

plt.plot([1,2,4,6,7,4,2,2], 'orange', linewidth=6)
plt.show()

Plotting expression data

f = open('../files/Spellman.csv')
header = f.readline().rstrip().split(',') # rstrip removes newline
data = []
for line in f:
    data.append(line.rstrip().split(','))
len(data)

for row in data[:10]: # first 10 only
    values = list(map(float, row[1:]))
    times = list(map(int, header[1:]))
    plt.plot(times,values,label=row[0])
plt.legend(loc='upper center', ncol=5, prop={'size':7})
plt.xlabel("Time");

Bar charts

plt.bar([1.1,2.1,3.1], [30,40,50], 0.5, yerr=[2,10,12], ecolor="black", alpha=0.5, label="One")
plt.bar([1.4,2.4,3.4], [60,50,20], 0.5, color='r', yerr=[10,4,1], ecolor="black", alpha=0.5, label="Two")
plt.xticks([1.4,2.4,3.4], ['A','B','C'])
plt.ylim([0,100])
plt.legend();

Histograms

import random
rdata = []
for i in range(1000):
    rdata.append(random.normalvariate(0,1))
    
plt.hist(rdata,10);

Combining chart types

Just accumulate plots

plt.hist(rdata,10)
plt.plot([-3,-1,0,1,2,3], [20,30,50,60,80,90], '-r', linewidth=3)
plt.show()

Terminology

Figure -- the full canvas you are creating your plot(s) on.

Axes -- a single plot

Multiple plots on a figure

subplots(nrows,ncols)

fig, axes = plt.subplots(2,1)
plt.sca(axes[0]) #set current axis to first
plt.plot(header[1:],data[0][1:])
plt.ylabel(data[0][0])

plt.sca(axes[1])
plt.plot(header[1:],data[1][1:],'g')
plt.ylabel(data[1][0])
plt.show()

See also GridSpec

Multiple plots on a figure

numgenes = 3
fig, axes = plt.subplots(numgenes,1,sharex=True,sharey=True)
colors = ['b','g','purple']
for i in range(numgenes):
    #most plot methods have equivalents on axes objects
    axes[i].plot(header[1:],data[i][1:],color=colors[i])
    axes[i].set_ylabel(data[i][0])

Saving figures

Instead of plt.show use plt.savefig

bbox_inches='tight' tells matplotlib to try harder to fit everything without extra white space

plt.axes(aspect=1)
plt.pie([8,2,1],labels=["A","B","C"])
plt.savefig("pie.png", bbox_inches='tight')

ls *png

Other formats: pdf, svg, eps...

Other chart types

Gallery

Seaborn

Making (some) hard things easy

Gallery

Activity

https://MSCBIO2025-2025.github.io/files/Spellman.csv

• plot expression over time for a gene
• plot the histogram of initial expression levels
• plot the histogram of final expression levels
• plot the above histograms on the same plot with transparent fill
• plot the above with identical bin sizes for both histograms
• plot the above histograms on different subplots, vertically stacked