Analyzing tabular data with pandas

featuring the human proteome

pandas

Pandas is the de-facto standard to read, analyze and visualize tabular data from CSV files, Excel tables and many more.

The main data structure for tables is called DataFrame:

Installing pandas

pip install pandas
pip install matplotlib

Your Task

1. Make sure you have the file human_proteome.csv.
2. Execute the paragraphs below in an interactive Python console.
3. Find out what each code segment does.

Exercise 1

import pandas as pd
import pylab as plt

df = pd.read_csv('human_proteome.csv', index_col=0)
df.head()

Exercise 2

df['name']
df[['name', 'length']]

Exercise 3

df.sum()

Exercise 4

df['W'].mean()
df['W'].std()

Exercise 5

df.describe()

Exercise 6

df[df['length'] > 5000]

Exercise 7

df.sort_values('length', ascending=False).head(10)

Exercise 8

df['H_percent'] = df['H'] / df['length']
df.sort_values('H_percent', ascending=False).head(20)

Exercise 9

def match_name(name, query):
    return query in name

hemo = df['name'].apply(match_name, args=["Hemoglobin"])
df[hemo]

Exercise 10

df.plot.scatter('K', 'R')
plt.axis([0, 1000, 0, 500])
plt.savefig('scatterplot.png')

Exercise 11

plt.figure()
short = df[df['length'] < 1000]
short['length'].hist(bins=50)
plt.savefig('lengths.png')

Where to learn more?

• on the pandas homepage
• in the book Python for Data Analysis