BSM 461
INTRODUCTION TO BIG DATA
Lecture 2 – Intro to Python
Kevser Ovaz Akpınar, PhD
kovaz.sakarya.edu.tr
kovaz@sakarya.edu
Agenda
• Python for data analysis and visualization
• Python basics
• Popular libraries
• Data manipulation
• Plotting
• Pandas
• Exercises
© Kevser Ovaz Akpınar 2
Python
• Very popular general-purpose programming language
• Used from introductory programming courses to production systems
• Software programmer Guido van Rossum from Netherlands in 1990
• Name is given from a show called Flying Circus by English comedy group Monty Python
• Its not scripting language!!
Python supports:
• Structural programming
• Object oriented programming
• Functional programming
© Kevser Ovaz Akpınar 3
Python Programming
• Many IDEs available or
• Notepad + Python interpreter or
• Anaconda which has Spyder and Jupyter Notebook software for Python programming
• Two versions of Python in use - Python 2 and Python 3
• Python 3 not backward-compatible with Python 2
• A lot of packages are available for Python 2
• Check version using the following command
$ python -- version
© Kevser Ovaz Akpınar 4
Python Features
• Dynamically typed
(rather than statically typed like Java or C/C++)
• Interpreted
(rather than compiled like Java or C/C++)
Python programs are comparatively…
+ Quicker to write
+ Shorter
+ Ease of programming
+ Minimizes the time to develop and maintain code
+ Modular and object-oriented
+ Large community of users
+ A large standard and user-contributed library
− More error-prone
− Interpreted and therefore slower than compiled languages
− Decentralized with packages
© Kevser Ovaz Akpınar 5
Python for Data Analytics
• Fairly easy to read/write/process data using standard features
• Plus special packages for…
• Numerical and statistical manipulations - numpy
• Visualization (“plotting”) - matplotlib
• Relational database like capabilities – pandas
• Machine learning - scikit-learn
• Network analysis - networkx
• Unstructured data – re, nltk, PIL
© Kevser Ovaz Akpınar 6
More on Python
• Reference types and Object cloning
• Most of the objects are Reference Type
• Functions are defined as “def” keyword
• Object oriented approach support
• “scikit-learn” library is developed in object oriented manner. It contains many
files like “naive_bayes.py”, which has classes.
© Kevser Ovaz Akpınar 7
Variable Types
• Numeric Types
• Strings
• Boolean Types
• Special Types
• Use the type function to determine variable type
>>type(log_file)
>>file
• Some keywords are reserved such as ‘and’, ‘assert’, ‘break’, ‘lambda’. A list of keywords are located at
https://docs.python.org/2.5/ref/keywords.html
© Kevser Ovaz Akpınar 8
Data Structures
• List (starts from 0)
• Negative indices allow access from tail to head
• List slicing
list[start_index:end_index:step]
step 1 as default
• remove() append()
• Dictionaries
• Stores (key,value). Key is unique. Dictionaries support add, delete and search.
• Tuple
© Kevser Ovaz Akpınar 9
More on Python
#lambda function 1
• Lambda functions fnc = lambda x : x + 1
lambda parameters : words print(fnc(1))
#Output: 2
print(fnc(fnc(1)))
#Output: 3
#lambda function 2
fnc2 = lambda x, y : x + y
print(fnc2(4,7))
#Output: 11
print(fnc2(4,fnc(1)))
#Output: 6
© Kevser Ovaz Akpınar 10
More on Python
• Easy: You can write a Python program in one single line into the Python
shell. So simple!
>>> def numpy_version () :
• Numpy api: Simple but not limited. Numpy: the main API used for what is t1 = time . time ()
called “scientific computing ecosystem.” Numpy handles linear algebra and X = arange (10000000)
matrix mathematics on a very large scale. Most machine learning Y = arange (10000000)
algorithms and neural networks operate on these n-dimensional matrices. Z = X + Y
✓ Written in C and Fortran return time . time () - t1
✓ Vectorized computations >>> numpy_version ()
0.059307098388671875
• Apache Spark has a Python shell. You can open datasets, do
transformations, and run algorithms in one easy command line. Without
that you would have to package your program and then submit it to Spark
using spark-submit. The disadvantage with spark-submit, as with any batch
job, is you cannot inspect variables in real time. So can print values to a log.
That’s OK for text, but when you use the Python shell that text is an object,
which means you can further work with it. It’s not a static non-entity.
© Kevser Ovaz Akpınar 11
More on Python - Matplotlib
• Used for generating 2D and 3D scientific plots
• Support for LaTeX
• Fine-grained control over every aspect
• Many output file formats including PNG, PDF, SVG, EPS
• Configuration file ‘matplotlibrc’ used to customize almost every aspect of plotting
• On Linux, it looks in .config/matplotlib/matplotlibrc
• On other platforms, it looks in .matplotlib/matplotlibrc
• Use ‘matplotlib.matplotlib fname()’ to determine from where the current matplotlibrc is loaded
• Customization options can be found at http://matplotlib.org/users/customizing.html
• Matplotlib is the entire library
• Pyplot - a module within Matplotlib that provides access to the underlying plotting library
• Pylab - a convenience module that combines the functionality of Pyplot with Numpy
© Kevser Ovaz Akpınar 12
More on Python
• The Python Pip Toolkit: Programmers contribute to its open source repository,
the Python Package Index (PIP). Sample pip packages read and write to JSON
and requests to work with web services.
• Pandas: Open-source library! Transform data from one format to another and
run these algorithms at scale, meaning across a cluster. For example, older
algorithms that existed before distributed computing (i.e., big data) like scikit-
learn would not work with distributed data frames and other objects run across
a cluster. They are designed to work with one file on one computer. So that is an
issue to keep in mind as you figure out which framework to use. With Pandas,
for very large data sets you might have a hybrid of tools
No support of parallel processing!!
© Kevser Ovaz Akpınar 13
More on Python: Pandas
© Kevser Ovaz Akpınar 14
More on Python: Pandas Comparison with SQL
SELECT total_bill, tip, smoker, time tips[['total_bill', 'tip', 'smoker', 'time']].head(5)
FROM tips
LIMIT 5;
SELECT * tips[tips['time'] == 'Dinner'].head(5)
FROM tips
WHERE time = 'Dinner'
LIMIT 5;
SELECT city, rank pd.concat([df1, df2])
FROM df1
UNION ALL
SELECT city, rank
FROM df2;
© Kevser Ovaz Akpınar 15
More on Python
• Python Notebooks (IPYTHON): Jupyter is used for notebooks. It is an interactive
computational environment, in which you can combine code execution, rich text,
mathematics, plots and rich media
© Kevser Ovaz Akpınar 16
Python on OS
• MacOS X, High Sierra has a preloaded version of Python 2.7 out-of-the-box. If you
have macOS X, you will not have to install or configure anything else in order to use
Python 2. If you want to use Python3, then installation is required
• Python doesn’t come prepackaged with Windows. Download the installer and follow
the wizard.
© Kevser Ovaz Akpınar 17
Python
© Kevser Ovaz Akpınar 18
Python
© Kevser Ovaz Akpınar 19
Python
© Kevser Ovaz Akpınar 20
Python
© Kevser Ovaz Akpınar 21
Python
© Kevser Ovaz Akpınar 22
Python -Spyder
© Kevser Ovaz Akpınar 23
Python – Anaconda Navigator other tools
© Kevser Ovaz Akpınar 24
Python – Samples
© Kevser Ovaz Akpınar 25
ADDITIONAL REFERENCES
Python and SQL Comparison,
https://pandas.pydata.org/pandas-docs/stable/getting_started/comparison/comparison_with_sql.html
Python ile Veri Biliminie Giriş,
https://medium.com/deep-learning-turkiye/python-ile-veri-bilimine-dal%C4%B1%C5%9F-3f069260ebda
Matplotlib Tutorials,
https://matplotlib.org/tutorials/introductory/pyplot.html
BYU, Big Data Science & Capstone Lecture Notes - Python
Stanford University Lecture Notes,
http://web.stanford.edu/class/cs102/lecturenotes/PythonData2.txt
Big Data Analytics in Python Programming, https://www.youtube.com/watch?v=G8VvTp0zgC0
Python for Big Data Analytics – 1, https://www.youtube.com/watch?reload=9&v=BiRXCLKLxrc
www.kaggle.com, “sf_salaries” Dataset
kovaz.sakarya.edu.tr
© Kevser Ovaz Akpınar
kovaz@sakarya.edu
