Digital Technology
Python – Functions, Pandas, MatPlotLib
Edoardo Ramalli
edoardo.ramalli@polimi.it
1
Functions
A function is a reusable piece of code.
The code inside a function is executed only when the function is called.
A function can accept some data to work with, called parameters.
A function can also return an expression to the callee function.
When the execution of a function is terminated the control flow continues, as
usual, after the call of the function.
Function - Definition
To define a function in Python we use the keyword def.
Then follows the name of the function.
After that you can define as many parameters you prefer (0 or more).
def function_name (parameter1, parameter2, …) # Remember the : at the end
# Function Body
# parameter1, parameter2, … are like variables in the body of the function
return <Expression>
A function define a scope variable, i.e. the variables that are defined inside
the function are not visible outside the function itself.
Function - Call
To call, to invoke a function:
# the result of the function is stored inside a variable
result_var = function_name(parameter1, parameter2, …)
# the result of the function is not stored
function_name(parameter1, parameter2, …)
Function – Default parameters
Sometimes a function can accept one or more parameters but some of them
could be optional. In case like this we can define default arguments with =
def add (num1, num2):
return num1 + num2
add (1, 5) à 6
def add (num1, num2=5):
return num1 + num2
add (1, 6) à7
add (1) à6
If you provide the optional parameter(s), the default ones are overwritten.
The default parameters can be defined only at the end of the parameter list.
Function – Multiple return
A function can returns 0, 1 or more expression. To return an expression we use
the keyword return followed by a variable number of expressions separated by ’,’.
Actually a function always returns a single data type when we want to return
multiple expressions, that is a tuple.
The tuple can store inside itself multiple expressions.
def my_function (…):
a=5
b = 10
…
…
return a, b
first_result, second_result = my_function (…) à first_result=5, second_result=10
print(type(my_function (…) )) à print <class 'tuple'>
Built-In
• print à print to console a variable(s)
• input à Collect a user input from console and returns a string data type
• len à Return the length of an iterable (list, set, dict, tuple)
• type à Return the type of a variable
• max à Return the biggest element from an iterable
• min à Return the smallest element from an iterable
• sum à Return the sum of elements inside an iterable
• round à Return the round of a number
• bool à Casting to bool
• int à Casting to int
• float à Casting to float
• dict à Casting to dict
• list à Casting to list
• set à Casting to set
• sorted à Return a new iterable that is sorted
Module
You can write many functions in a file (called module) and use them in another
script by importing them. If you have a function called ‘average’ in a file
‘myfunctions.py’, you can use it in another script with:
import myfunctions
print(myfunctions.average([1,2,3]))
Module
Python comes with a standard library, offering a wide variety of facilities:
https://docs.python.org/3/library/
For instance, we have the math module for the sqrt function:
import math
print(math.sqrt(9))
For statistics stuff (avg, median, stdev, etc..) instead we have:
import statistics
print(statistics.stdev([1,2,3,4]))
We will add other modules to python with pip and https://pypi.org/.
Install others moduls
We have seen how to import a module from the Python standard library:
import math
from random import uniform
However, there exists many modules which are not included in the standard
library (numpy, pandas, scipy, scikit-learn, etc...)
You can install, upgrade, and remove packages using a program called pip:
pip install pandas
pip uninstall pandas
pip freeze
Pandas
Pandas is a data analysis and manipulation library.
Pandas allows you to open tabular data (like csv, spreadsheet and databases) and
obtain Python objects called DataFrame.
https://pandas.pydata.org/docs/getting_started/index.html
Pandas - DataFrame
Let’s manually create a DataFrame:
# Instead of using every time ‘pandas’ you can define an abbreviation with the keyword ‘as’
import pandas as pd
my_df = pd.DataFrame(
{"Name": ["Alice", "Bob", "Charlie", "Tom"],
"Age": [25, 30, 35, 27] })
print(my_df)
Each column is a pandas Series.
You can select a column using square brackets:
print(my_df['Age'])
Pandas – DataFrame filtering
Let’s keep only the rows of dataframe table when the age >= 30
print(my_df[my_df["Age"] >= 30])
Now we would like to add a column in the dataframe.
This column reports the age of the people present in thedataframe, in 50 years.
my_df["Age_50"] = my_df["Age"] + 50 # The new column is Age_50
The new column is built in this way: for each row, we take the corresponding
value under the column Age, and we add 50 to the value and store the result
under the column Age_50 in the corresponding row.
In similar way we can perform operations between the column of the dataframe.
print(my_df["Age_50"] - my_df["Age"])
Pandas – Import/Export a Dataframe
You can save a DataFrame to any of the format supported by Pandas:
import pandas as pd
df = pd.read_csv("apple.csv")
df.to_excel("./apple.xslsx")
We will see more on pandas in the next lessons
MatPlotLib
Matplotlib is a comprehensive library for creating static, animated,
and interactive visualizations in Python.
The most used module of Matplotlib is PyPlot, which makes Matplotlib work like
MATLAB. Each pyplot function makes some change to a figure,
and various states are preserved across function calls, so that it keeps track of
things like the current figure and plotting area. (=> Not a RESTful API)
To install MatPlotLib:
pip install matplotlib
https://matplotlib.org/stable/users/installing.html#
https://matplotlib.org/stable/gallery/index.html
MatPlotLib
If you provide a single list or array to the plot() command,
matplotlib assumes it is a sequence of y values, and automatically generates
the x values for you.
MatPlotLib
Homework
Write a function such that accepts a dataframe of a stock with columns:
Date, Open, High, Low, Close, Volume, Dividends, Stock Splits.
Each row of the dataframe represents the opening price, the closing
price, etc., of a specific stock in a specific date reported under the column ‘Date’.
The function should returns the biggest rise date an the biggest drop date, in
percentage variation, between all the stock history available.
The csv containing the data information is available on Beep in the Python Class 3
folder.
!"#$%&'
Percentage variation = %&'
∗ 100
It is not mandatory. It is not graded.
