Python data science

A Comprehensive Guide to Self-Directed Python Programming Learning

Vere Salazar
Copyright © 2024 all rights reserved.

No portion of this book may be reproduced in any form without written

permission from the publisher or author, except as permitted by u.s.
Copyright law.

This publication is designed to provide accurate and authoritative

information in regard to the subject matter covered. It is sold with the
understanding that neither the author nor the publisher is engaged in
rendering legal, investment, accounting or other professional services.
While the publisher and author have used their best efforts in preparing
this book, they make no representations or warranties with respect to the
accuracy or completeness of the contents of this book and specifically
disclaim any implied warranties of merchantability or fitness for a
particular purpose. No warranty may be created or extended by sales
representatives or written sales materials. The advice and strategies
contained herein may not be suitable for your situation. You should
consult with a professional when appropriate. Neither the publisher nor
the author shall be liable for any loss of profit or any other commercial
damages, including but not limited to special, incidental, consequential,
personal, or other damages.
 Table of contents

Introduction

Chapter 1: what is the difference between data science and analysis

Chapter 2: numpy introduction

Chapter 3: manipulation data with pandas

Chapter 4: visualization with matplotlib

Chapter 5: machine learning

Conclusion
 Introduction

of it is to you of data also providing you with all skills and tools to support
your this book is for looking for a way to basics of data
data with python can intimidating at to of programming and
can to can along string of that by you will basic tools that can you charts
and it for you to your with
in mind that to most from this book, you should work through in if you a
difficult of apart by and slowly push through until you will you about
tools you show you how to install and you information so that you can
work on your own as to you without practical work, you won’t at fully

a lot of options that can work with it to data and almost all going to to this
kind of and of if you wanting to know about your and what looking for
from you and your data is going to option to you out.
To start this going to a look at what data is all about, why it is important,
and why would want to work with this in first as will to of data and how to
go through a of finding right data, data, coming up with right and
going to taking a look at basics that with python and how can this to with
our data a lot of coding out that can work with it is to data but many that
and to and of python it of to with this will also looking at a of that going
to work with it to python, including numpy and its arrays, and matplotlib
to all of work in no
This is just start of of fantastic things that can do it is to start on data can
our looking at what is all about, of and how can put it all it is to sort
through our data and find right and insights in
a lot of that can it to working on data many in a lot of going to work with
this to that will to how to work with how to out and so much you to work
with of data and you want to work with all of parts that found with it, to
out this to you
 1: what is data and analysis

What is data

Data is a that you about at a if you running a company, and you a look at of
that you can to and to risks. it is to big and important you most about data
from who in boat as you. but this that much about data and what can do with
it. and that is this is going to into play.
As world to of big data, for right to with it was that to grow crazy as too. this
was of main and that with at until a ago. main focus of this was that might to
and a lot of data for didn’t an to all of that data to work with
Now, thanks to of that in that this is no big that it was in past. and this that
can a and work with and data in and you will find that that is going to bring it
all and us not only to up data that but will that can what is found in all of that
data is going to data

a lot of that looking to into world of data know all of that with this as to of
data that is going on with all of this, to that right data, and that all of
information that with at but that is all going to in data analysis that doing
along way.

Why is data so important?

first thing that to a look at doing of our work is why data is so important, and
why a company may to work with of data along way. first is way that this
data has in our world, it is to how it was in past. traditionally, data that to
hands on was going to in and it was much this us to work with tools to what is
going on.
this is not any and with this, of it allows us option to know about a situation
can up data and with data. but this data is going to and that it to sort through
and as

of data that was found in traditional that and to work with, today, will find
that most of our data is or at this is going to it to work with and can but can
find information to that making, this is not a bad thing all of
This data is going to from a lot of including things forms, and financial logs
along way. of all that data from, it is to that not to work with of tools not of
handling all data. this is why it is important to work with data to work with
algorithms and analytical tools that and as this that can and insights out of
data as
on data that car can it can to it it should up, go it should or to turn with of data
and to of algorithms.

And can work with of data to out with analytics. for can work with data from
things radars, aircraft, ships, and can to us build up of that going to can also
to of any natural that would it is also going to us to of right of this, and can, in
us a lot of as

How can i data

of first things that will it to data is of how can it. it may a thing to about and
can work with it just for fun, but a practical that going to with using data no
what industry you in or how you plan to it. of ways that can on data will

It can with is all about data, and can all of big data that is for data will find
that can about our and how can to a ways that a company will to data to of
and it can with knowing how to placing in right for right and
company to about as a company, that you can about your off you will about
what your want out of your your and how you can will that will to pick out
your than working with along way. it can you to which products you going to
It can to all want to about and amount of that is going ton in company. this
can parts, and part of of making product. can cut down on of this it can
company to and will profits that company is going to you can data you will
find that it is to cut out of this and that you would

Financial institutions it to fight off fraud. fraud can cost financial institutions
a lot of if not with how and don’t how to watch out for transactions and going
to along way. financial institution can up with algorithms and programs that
can and stop fraud, it can up saving millions of dollars a
This is data can in, as using a lot of factors, as as training algorithms on past
transactions that it is for of to out a transaction is or not. and that algorithm
through it is going to at doing that work, and of fraud that going to
It can up loan applications. a amount of loan applications that into banks and
financial institutions today. this is a lot for a loan to and of financial
institution to do on own. data can into this to

With of data financial institution can go through applications and out which
loans and not. you can up a of that application to and work from if application
of it is going to a good option to work with. loan will and can final on it. but
if loan application it will and not at all. this to up loan application and can
allow loan a to up with work and focus on who will bring that financial
institution
It out in in a of ways. so many ways that of data would to out to start with,
can that data is going to out with doctors and finishing up with of and
algorithms that with data doctors going to to right incisions and with minimal
and with
In addition, algorithms going to doctors a prognosis about with workload that
is on doctors and to with having can will to data to look through of that such
as scans and x-rays, and a diagnosis in no at all.
And finally, can find data in work it to tasks that in can talk to ask a and out
of of and can to right location. for that running low on staff to of this can

It a company about and industry in. if you want to out of that you right now,
you to find out as much information about your industry and your you to what
doing, how to it, and how you can this to your
you all of this information in you will find that it is to out what is working
and what is not working. you can jump right in and of to your you may find a
way to your that no is working with. you may to modifications to a that is
working with that can your own as
Many manufacturing going to work with data it allows to that can first, you
can kinds of algorithms to out a part on a is to or you to do you can this to
during off-hours, that is always up and running on and that production is not
going to you things out.

algorithms can also work, in much any it is to work with finding ways to cut
down on of that is company is going to kind of and figuring out how to that
and of it, will it so that you can products at a still making as much in profits
as it is as as working with data to out and how you can to a minimum.
you doing shopping, and you that is a tab on that of work of data and coming
to so popular on many can it for to on what want to watch or or and going to
as with it is that is going to which is going to bottom for that company.
on list is going to of if you and a to your or you with of popular cortana, and
you can how this is going to fun and and it is run with of of data
In fact, of that going to find it to data is this using if you not in a position to
out a you still about to talk to or and ask a this may not in this that may to
working with program and it how to of things that saying. and this is that is
through and will to how to of and

If you with of that out that out with comparisons, you data at work as at a
basic kind of you will find that that will with of lots and lots of data that can
with of rss and apis. can this information to out what of product from and will
so you can and this can in many to you and out what is for you.
can also a look at how this kind of is going to work talk about planning. this
industry is that is going to a lot of if not for a many struggling to maintain
occupancy that and profits. and add in of right now, and to discounts to many
situation is slowly
It wasn’t too long to work with data to how can profits and without scaring
away. this to out of that can today, with of data can:
about a flight
which class of should in.
should land at or a halt in
to of loyalty programs that can sign up for.
Two of who will and and now of most of out

And finally, going to look at logistics to that in and that things don’t lost. ups,
and dhl going to data to how
With of data and to out to ship, that should and of transportation that is for as
this to cost and along way as in addition, data that can with gps units can with
data to and to with data along way.
so many that can it to using data for a company. in many if you not willing to
data at a bit, you will find that you will fall no what kind of you trying to
work on or what you would to in your you will find that working with data is
of and most to you out.

Data analysis

on topic of working with data also to a look at of data analysis. ]this is

basically a of data and it is going to will data that which in from many and
using algorithms to us out what all of this data
You can all of data that want, and can fill up a lot of data if would but if you
don’t what data is us, data is not going to all that much, and your company
will not to it. this is will to grab of data analysis and it into play for us.
To us data analytics is going to or of taking raw data that and analyzing it to
us good conclusions about all of that information. many and with this into
algorithms and that can work data in its raw form turning it into that can
through and than
data analytics it is going to to us of and that may miss out on in of situations
out is just so much data that and holding onto. this information is going to
what can to to how a or a can
Now that know a bit about broad of data analytics, it is to into of parts of it,
and how can it to our data analytics is going to broad, and it many of data
analysis. any of information that can to of data analytics to that can by
humans to about will fall this
look at an of this. many manufacturing going to work and run for that
working with. will data that contains this information, and it to plan how and
workloads should this as to capacity as saving and in

This is just of tasks that data analytics is going to to out with, though in
production is a good thing to pay to. gaming going to to this to up right for
which in a way that majority of stay in than dropping off and using it. is a
company. can various parts of data analytics to clicking, watching, or so that
can clicks and that
As working through this always to in mind that data analysis that working on
is a and not that can just at and up with that want. if you do this, it is that you
will up missing out on important information, you work with of of data.
Following right along way can a big in what going to with this data analysis.
and of that can will following:
start this by of data or how data has a lot of including data by things and of
data could also or may up by as

in of this is of our data. a ton of you can all of this information. it could from
your from in and social and from your
data has and you you data that you it is to data you it from a of you to that it
is in a that and that algorithm will to through quickly and find to from you to
that
You can what you would to you information. you can work with a or try kind
of that is good at handling data that is statistical.
you to data way that you would it is to it all up analysis this that data has to
go through and scrub data and that no or duplications that found in
information, or that data is not this but it that is and to go you start.

thing that to is that it is going to that can to raw data, it, and good conclusions
on this information as and that in this kind of thing is going to in many and
you will that work with right algorithm, it will to all of that raw data and turn
it into that humans can and main that a would want to this kind of and to data,
is it is going to so it is to your your and how you can your

Why analysis of data

to a to look at why this data analysis is going to so important, and why a

would to go through algorithms and to from that data to in first find that can
insights and that in data, but is such a amount of data, and all of that data to
from a of it is going to a lot of work to sort through—too much work to try
and do manually.

that of data analytics is that can to any no what industry in it will to it can also
to work on and can it for to that for costs and things out.
In addition to all of this, a company will to that data analysis and that going to
into this analysis is also going to to about of that out and can you to and
products that going to work for your as

in data analysis

thing that to of our is to out what kinds of analysis for you to work with as a
lot of variations that with data analysis, and you can pick that works for your
a of but going to focus on of that you most to you first on this four basic that
a in data analysis can focus on, and that most data going to work with will
following:
analytics. this is that will to what has in data a you may look at things of has
up, or if for your company or than in month for a product.

Diagnostic analytics this is that will focus on why has or root of how your is
doing. this is but it will data inputs that and a bit of in for did you a slow was
and no out shopping? or did go wrong with of your campaigns, and that is
why struggling?
analytics. this is going to to what is most to soon, on information that at our
disposal right now. what to last was warm? how many going to us that will a
hot this and should for of what saw in past?
As can show up with this, though going to focus on handling data and us to of
insights that from data. to this data and in an and through analysis is going to
though not to all of this work

without a with this, can how to data analysis work and can this to a ton of
data and from it. all of of doing this, and that this is going to to smart such by
data so that going to to a lot of and how to out

What of a data analysis?

it is to work with data analysis, it will not you too long to that a lot of that
going to with this as no what kind of company you in and what your goals
you will to a lot of in as this may it has a ton of and that you going to too long
working on it for it to worth your but it can so worth it if you to about data
and all of cool things that it can do for you. with this in mind, of that you
going to it is to work with data analysis for your company, no what kind of
company, will
It you to your all want to this is way to and but how you to know what wants,
and what is going to to to your to going to a this data analysis can information
from and habits and you to to as

It you to know what going on that you should follow for your always that go
on in any but shifting and changing at fast that hard to up with. using a data
analysis can you to catch on to of of making it for you to of your
It you to know your product this data analysis can to you know which
products doing and why. you may find out that product is doing than you
thought, or that you should start of a similar product to
It can it is always if you can data and information all of that you for your
company. data analysis you to comb through all information and what is you
any about your company.
This data analysis is going to of ways that you can out who willing to look
through data and willing to what insights and who will to find things than and
who will to win in industry.
You will find that going to out your company, and of main why this data
analysis can so good for company, and why it is that is in such high for
almost industry that is out for of that out who may not choosing to work with
this kind of will find that can still work with it on a local and as
Data analysis has and to out how to this and what you can do to of it to what
is in that data, is going to important as to go through of that about in this in of
this and you will that you can this work for your as
 2: numpy introduction

array

You can build data with at storing data; not at on that data. optimal it to
and which critical working with algorithms. this is why using numpy and
its ndarray which stands for an array”. look at of a numpy array:
It is optimal and fast at data. you work with data, you want to it of
You can in words, you can computations and without to “for” loops. this is
a plus for numpy python “for” loops cost a lot of making it to work with a
of loops of ndarrays.
In data you will to tools, or such as scipy and you can’t without arrays as
an input, functions won’t as
With that said, a of a ndarray:
an data and turn it into an array.
Build array from start and add in

You can also upload data to an array it’s on a disk.

a list to a array is a fairly common in data in mind that you to into account
of such a list contains. this will an impact on of an of this with a list that
contains only

In: import numpy as np

Int_list = [1,2,3]
Array_1 = np.array(int_list)
In: array_1
Out: array([1, 2, 3])
You can array just you a list in python. you simply and just in python, it
starts from 0. this is how this would look: in: array_1[1]
Out: 2
Now you can gain data about array so:
In:
Out: numpy.ndarray
In:
Out:
of is to of running. in this using a 64 bit
how can load an array from a arrays can from data a an in
In: import numpy as np

Cars =
In this our tool to an array from a with of “loadtxt” by giving it a and a
data

of numpy arrays
thing that you will quickly it to work with that to work with python is that
many of going to on numpy array. this that will to at a bit of about arrays
and how can work for of our as arrays going to to us in many of parts that
want to do with data so how to work, and what can do is going to a big
first thing that to a look at is what numpy array is all about. going to work
with what is known as a numpy array. this is going to an array of which
going to similar to what with c will also that arrays going to which that
only going to to contain data that is so of an array that is an or a floating
can go through and of arrays with as
Now, if you working with python you may look at of arrays and how
going to any than what you can with python list. going to a for this, and
going to a bit of to gain a of what is going to with both array and list

To start with, array in numpy is going to compact than what going to with
lists in python. a list of python is going to about 20 mb or of data to and
this can go up if you working with a or list on but if you working with a
3d array in numpy and has a float in a it is going to about 4 mb. this can us
to to information a lot and you working with data analysis or a data that
has a ton of information, of your coding parts will off for you.
Also, you would also find that array list in of and writing is going to a lot
and to you to work with array. this will you to through of arrays that you
would to a bit and will that you going to things you at of data to through,
and you working with you will find that all of this can in long run.
Now, in it may not in or that with python lists and numpy arrays would
that big of a but you will find that a ton of this is going to important to
what you trying to with of your

It may not that much if you just working with a million but it is going to
you working with a lot a billion approach is going to fit into that is on 32-
bit. still, if you working with a 64-bit, you can build with numpy and it all
with about 4 gb of data python it is doing all of works on its own, would a
minimum of 12 gb to it and many will say that it as much as this if not and
can that that to will to on to of it.
Not only is numpy going to about but you will also find that it is a lot and
to work with. you work with this you going to find that you can a lot of
matrix and options to work for you for this is going to it allows you to
avoid of work that is but which you to do to of parts of python, lists to
work. add into this that you can and it is no that so many want to work
with arrays for of data analysis
To it a bit not only you going to that with arrays, but you will how along
with functionality and it to sorting through millions and billions of data
points it is that can do this in a fast and along way. this is that numpy array
is going to to us with.

first thing to look at is going to functionality. you going to to build up a lot

of things you working with numpy array. for you can work with things of
basic statistics, fast ffts, convolutions, and histograms, to a and how you to
to of data that you want to do without of important parts showing up as
And thing that you will to and it to working with numpy array is that it has
a lot of may not that you working on a that a lot of data points to it. but
you working with a big with a lot of data points, you will quickly find that
this is going to important and that numpy array can on.
of that you will to you to work with numpy array than python list with of
your data analysis will of following:
arrays that show up in numpy going to out at a at this is than what can it to
python list that can and grow dynamically. changing of your array it is is
going just to a array that you can work with.
that you going to with of arrays going to us that must kind of data can
work with as but this it to do of that you want to do. and as this is going to
it so that you can’t do of parts of data that you want with arrays, it is not
usually you using of in all of this, you will find that all of going to in of
that you writing.
arrays that working with going to and of it is to work on of data. typically,
going to and with than what can with of built-in that going to with python
as
final thing that you going to it is to work with numpy library is that many
of that work with python, including for data and data analysis, going to on
numpy arrays. it is hard to do any of functions that you topics if you do
not numpy array in first, and to do this. to that this library up and to go as
As to a ton of and that going to to it is to work with of numpy arrays. to a
lot of options it to handling our data and how to work right way is going
to critical to how much going to to with this that you want to work with
kind of data that you to install and important numpy library. this is going
to that can out of and will it for us to that would is

Numpy array

you work with arrays, you may want to add rows or columns to data and
this is known as array stacking, and it long for numpy to information. start
by a array: in: import numpy as np
=
add a row, and that will
In: =
=
Now vstack which stands for a stack, to add a
In:
This command will also work if want to add
In:
how to add a to array. this is with “hstack” which stands for horizontal
stack. an in: bias =
In this of bias of unit to array
As an aspiring data you will only to know how to add rows and columns
to your arrays. in most you won’t to do than that, so working with arrays
and numpy this tool is in data

What is numpy library

Now it is for us to a look at of that can work with it to using python and
our data analysis to work for our numpy is of first that can look at, and it
is going to of it is going to basis that can with of important that will
discuss on, or data analysis so it is worth our to a look at it.
To start with, numpy is a library that is in python. can it for including as as
computing if it. for most part, though, it is going to to us our array s
quickly and will it and out in python and c
though this is a that works for c as this is going to a basic data analysis
library that going to with python. word numpy is going to stand for
python. going to bring out this library to us to any of arrays that want to
This library is going to of that is for computations. this that it is going to a
array of and it will tools that it is to work with arrays as

You will quickly find that you work with data analysis that talking about
that numpy is going to in almost all of programming that try to do with
python, including things statistics, and bioinformatics. it is also going to
us with good functionality that can work with, functionality that can work
will run and is in

about numpy

This library is basic, but it is still going to important it to handling of

computing that want to do with python. plus, it will not that long working
with data and data analysis you find that this is going to library that data
analysis going to on.
of significant going to on arrays in numpy as inputs and outputs. in
addition to this, it is also going to functions that going to allow a way for
to all of basic and functions that would talking about statistics or with
arrays and without to as many of to it all
arrays with lists that about with python, you will find that arrays going to
much but python lists do an arrays as you only to data in column working
with arrays.
a that you going to it is to work with numpy library. of main that you will
most will

numpy library is going to a combination of python and c

This is going to consist of arrays that and ndarray is part of this as which
will arrays as
It is going to work on a lot of functions for arrays if you would
It can also us to arrays. it also allows python to a way to work as an to
matlab.
a lot of why would want to work with numpy than having to pick of that
out along way. will array in numpy for work that doing with python of a
list. and of for this it is to work with, it is going to than and it is going to
All of going to important trying to do of algorithms that in data analysis.
and mostly, you will that arrays going to thing that it is to work with this
library as

a things that to it to how can work with numpy library. first, numpy array
is going to up a lot than options. on hand, can an array, and it is only going
to about 4 mb. if you to a lot of arrays as you go through, and going to fit
on of your arrays also going to to you would to and on on.
Also, it to you will find that numpy arrays going to it is going to to a lot it
to computations than what find with python lists. this library is it is not
going to cost you anything to with. it also has of working with popular
python programming which has high-quality for almost all of tasks that
you want to accomplish.
All of to work with. you will find that it is a high-quality library that is
going to us to things you can it to match up with that you want, it is going
to to work with, and it can a lot of data analysis that you want to do. it is
also an library that will of that in c to for python so you can your work
a lot of that going to up you want to work with numpy library, and you
will find that it is going to basis for a lot of and algorithms that you want
to out you working with your data analysis. how to this and what it can do
for you is going to a world of in how much you can accomplish for and it
is worth your to about it as to your
 3: manipulation data with pandas

Pandas

Pandas is built on numpy, and to this it to arrays from data arrays can into
data a look at an
In: import pandas as pd
Import numpy as np
=
=
In this uploading data to a data going to to an array that is of as data
In: _array =
Out:
can that a float array. you can of array by first using this will which by
data do this array. this is how this would look: in:
Out: 0float64
1 int64
2float64
3 int64

4float64
5float64
6float64
7float64
8int64
9int64
10int64
11float64
12float64
13float64

Matrix
This matrix calculations, such as matrix to matrix multiplication. a array.
This is a array of from 0 to 24. will a of and a column that will stack and
its what it would look in: = np.array([1., 0.5, 0.5, 0.5, 0.5])
=
Print
Out:
[[1. 0.5]
[0.50.5]
[0.50.5]
[0.50.5]
[0.51.]]
Now can multiplication. an of multiplying array with
In:

Out: array ([5.,20.,35.,50.,65.])

an of multiplication array and
In:
Out:array([[5.,7.],
[20.,22.],
[35.,37.],
[50.,52.],
[65.,67.]])
In both of multiplication “np.dot” function in to up, discuss slicing and

Slicing and

is for ndarray by an instruction to of columns and rows or

start by a 10x10 array. it will initially
In: import numpy as np
M=
rows from 2 to 8, but only that
In: m[2:9:2,:]
Out:array([[20, 21, 22, 23, 24, 25, 26, 27, 28, 29],
[40, 41, 42, 43, 44, 45, 46, 47, 48, 49],
[60, 61, 62, 63, 64, 65, 66, 67, 68, 69],
[80, 81, 82, 83, 84, 85, 86, 87, 88, 89]])
Now column, but only from 5.
In: m[2:9:2,5:]

Out:array([[25, 26, 27, 28, 29],

[45, 46, 47, 48, 49],
[65, 66, 67, 68, 69],
[85, 86, 87, 88, 89]])
rows and columns. but what if try a doing so would array. how our array
would look using a
In: m[2:9:2,5::-1]
Out:array([[25, 24, 23, 22, 21, 20],
[45, 44, 43, 42, 41, 40],
[65, 64, 63, 62, 61, 60],
[85, 84, 83, 82, 81, 80]])
ways of slicing and arrays, but for this book, it’s to know how to in mind
that this is only a way of data. if you want to by data, you cannot any
modifications to original arrays. if you do, it can to in that you want to
“copy” this will a copy of array which you can modify you wish. for copy
in: n = m[2:9:2,5:]. copy()
Pandas built on numpy, and to this it to arrays from data arrays can into
data a look at an
In: import pandas as pd

Import numpy as np
=
=
In this uploading data to a data going to to an array that is of as data
In: =
Out:
can that a float array. you can of array by first using this will which by
data do this array. this is how this would look:
In:
Out: 0float64
1int64
2float64
3int64
4float64
5float64
6float64
7float64
8int64
9int64
10int64
11float64

12float64
13float64

Matrix

This matrix calculations, such as matrix to matrix multiplication. a array.

This is a array of from 0 to 24. will a of and a column that will stack and
its what it would look
In: = np.array([1., 0.5, 0.5, 0.5, 0.5])
=
Print
Out:
[[1. 0.5]
[0.50.5]
[0.50.5]
[0.50.5]
[0.51.]]
Now can multiplication. an of multiplying array with
In:
Out: array([5.,20.,35.,50.,65.])
an of multiplication array and
In:
Out: array([[5.,7.],
[20.,22.],
[35.,37.],

[50.,52.],
[65.,67.]])
In both of multiplication “np.dot” function in to up, discuss slicing and
Slicing and

is for nd-array by an instruction to of columns and rows or

start by a 10x10 array. it will initially
In: import numpy as np
M=
rows from 2 to 8, but only that
In: m[2:9:2,:]
Out: array([[20, 21, 22, 23, 24, 25, 26, 27, 28, 29],
[40, 41, 42, 43, 44, 45, 46, 47, 48, 49],
[60, 61, 62, 63, 64, 65, 66, 67, 68, 69],
[80, 81, 82, 83, 84, 85, 86, 87, 88, 89]])
Now column, but only from 5.
In: m[2:9:2,5:]
Out: array([[25, 26, 27, 28, 29],
[45, 46, 47, 48, 49],
[65, 66, 67, 68, 69],
[85, 86, 87, 88, 89]])

rows and columns. but what if try a doing so would array. how our array
would look using a
In: m[2:9:2,5::-1]
Out: array([[25, 24, 23, 22, 21, 20],
[45, 44, 43, 42, 41, 40],
[65, 64, 63, 62, 61, 60],
[85, 84, 83, 82, 81, 80]])
in mind that this is only a way of data. if you want to by data, you cannot
any modifications to original arrays. if you do, it can to in that you want to
“copy” this will a copy of array, which you can modify you wish. for copy
In: n = m[2:9:2,5:].copy()

Data munging with pandas

Now that up with all tools you and with a it’s to with, data is known as
data munging. known as data wrangling, this is of most important in data
basic it is that you to a of data to to it with of data or to it. you will to an
original to it for your goals. this is a
To this that you a of data on which you to apply a classification algorithm.
you that you can’t this just is a combination of continuous and this that
you to modify of to match format. is that you with raw data, and you
cannot it just first, you to data with various data munging and tools.
In most you will with a of data that is raw and cannot just for study and
this is why you to data, and it can you a of many data data of coding or
running various algorithms. so how do our data? of most popular using
pandas library, which is for data analysis and manipulation, as of this
library, in this is to allow you to raw, data a lot

in mind that of data munging is to information to to within it. data to to to

a data or analyst to cut down on and to up with with that in mind, first you
to acquiring data. all, without data, you can’t do anything. you gain to it,
you to that all data and not you will data with an is about joining data it
has from at this data to and to at a with analysis. finally, last that data. this
is main you will to modify data to obtain a format you can you might also
to optional noisy data or bad data that can
As you can this is although and you cannot avoid it. you will to that you
up to information that contain any null to only data you in for analysis.
python, with pandas, of most tools you can to aid you in data munging.

You can data in many ways. you can it from a from from and any digital
that holds information. this is your raw data, you cannot it for a analysis
among all that data, you also missing information and data. you to first
bring to chaos by using python data to turn raw information into an data
out of this is with of various algorithms.
you can your data to up with an that you will you will obtain by you and
up data with various such as graph analysis and most you will now to first
data your will you that you will to apply any to your data and to data
munging to in mind that, in most output you will not output that you will
always you in you and that is why you to many and to what works.

Importing

can do anything, actual data. to import a going to pandas to tabular

information from various or this tool will build a data in which row of
tabular will also so that can data. with that said, will work in and
following command to import pandas into our and a csv
In: import pandas as pd
=
Iris =
iris in 2019)

As you can first of this is to import tool going to you going to pandas or it
is not to just it on your you to import it into your to to its functions and a
and it, also that will act as a and a in this will contain an that has to data
and for is iris, and it contains 50 of of iris also that don’t want to a it is not
in our what did so far was to a data iris, which is a data discuss it in of
working with pandas. in this a data is as a python list or dictionary, but
with a of want to what data contains by typing:
In:
This is a instruction without any by if don’t going to first rows from if you
want you simply to of rows you want to by typing it as an of function.
want to of columns to what kind of information contain:
In: iris.columns
Out:
As you can in output, what for now is an of column of output looks a list.
now, obtain column:
In: y =
Y
This is what you should as output.
Out:
0
1 iris
2 iris
3 iris
…
149 iris-virginica

“y” in this is a typical of pandas. what you should know is that it is to an

array. it is only you will that class is as for a dictionary. going to to list of
columns by using
In: x =
Now data which is a matrix of a why it is a matrix is that to columns at
and mainly an array that is in columns and rows. now also obtain all of
In: print
Out: (150, 2)
In: print

Out: (150)
is now a and can of array in basics of manipulating a and basic on to and
data so that can it.

Data
Now that you know how to load a you to to all information within it. first,
going to that to a action on rows. to any function, first to up a mask. that
in this a mask is a of that practical will up this notion, so to it:
In: = > 7.0
In:
0
1
…
146
147
148
149
As you can only which contain a that is than with a going to apply a mask
to modify iris virginica and a for it:

In: = == ‘iris-virginica’
In: =
old iris virginica it will now with as that for this to “loc” function to gain
to data with of now and that by column:
In:
Out:
Now group all of columns:
In: =
Out:
In: =

With this columns by using group by function. that this is similar to

“group by” command that you in sql. in input also function, which of in
mind that can apply this to a column or columns at can count, or sum
functions to gain that you obtain is also a pandas data which you can all of
in this grouping all data by to to all groups. but what if also to with?
In you familiar with you should know that imply analysis of a of data that
in this is chronologically. in you with a group of points that in with an
dividing of you will in many usually statistical analysis. for you to work
with data, you will find a or of sunspots.
is with data that contain in mind that training and for to and in fact, and in
world, you will with noisy data. in that first thing can do is a rolling
function, which looks this:
In: = 5)
For this applying function in mind that you don’t to you can also go with
in addition, you will that only to going to apply function to on our
columns and rows. this is a function that can for so start by how many

In: iris.apply
Out: 0 5
15
25
35
45
int64
Finally, applymap function is to on say that to obtain of for string
In: iris.applymap (lambda
In to obtain to a string, and can
Now that you an about using pandas for data also discuss topic of data
with of tool.

Data

What do to data that you a with an column that you to to modify it and
work with it. for of this going to that starts from 100, so:
N,val1,val2,val3
100,10,10,c
101,10,20,c
102,10,30,b
103,10,40,b
104,10,50,a
As you can first row is row 0, and its is 100. you import you will an
column, as usual; possibility of changing it or using it by it would a good
to split column from of data to avoid making any you running low on
going to pandas to column and it apart from
In: =
That’s it! now you can as usual, you by column, or locations. for you can
fourth from fifth which in our has an to 105. how looks:
In:
You might to this to a matrix; it isn’t, so not to confusion. in addition, you
should always column you want to you row. this way, you won’t any
looking to gain to a
your data and information about it can you along so always to for data
munging and in going to going to gain insight into data and data as a
 4: visualization with matplotlib

How to work with matplotlib to visuals

library that can work with, working on our own visuals as about is
matplotlib library. this is going to a option to work with of work that want
to do with a data and will that can all of data points, and all of that to with
our algorithms and of that from to us with visuals so that data a bit
To start with, matplotlib is going to of plotting that is to work along with
python programming it is also going to of that with numpy library that
about big data, and of handling matplotlib an api that is to plots in
applications of python as
python is going to in in many numpy and matplotlib going to in many to
out with of that to do with and to that can work with

that going to with this is that can to of tasks that of that with python. it is
to go through and with all of and of program for a and of options that
would to
You will find that of utility that can with this library, as as with numpy is
going to around utility of matplotlib is going to with visual plotting tools.
so, in a going t analytical than all of that can with this library is going to
allow for programs of can us right for human as
With of this information in mind, to look a bit about matplotlib library in
to start with, this is going to part of from python to with 2d graphics. how
to work with this kind of library is going to so important you would to of
visuals and that you want to do in a data
What is matplotlib?

This is going to of plotting that you can in python, you would to things 2d
graphics. it can with a lot of on application python python script, and of
toolkits out that graphical

going to toolkits that that will to out of functionality that going to with
matplotlib and will that can do with this program in no at all. of going to
us going through a download, and going to found with of this library but
will to on a that not found in python or this library. of that can focus on
and can work with it is to out what matplotlib can do will
this is going to a map plotting toolkit that can if this is what you would to
work with of your it is a good option to if you would to work with
political and map
Natgrid: this is going to an that to natgrid library. this is want to gridding
of data that
Mplot3d: this is going to you would to out 2d functions of matplotlib into
that is 3d in
tools; this library is going to us with of that to data with microsoft if it.
Cartopy: this is going to of mapping that can work with that going to us
with of of map and of arbitrary point, polygon, and transformation to a of
that can on.

a lot of options that can work with along way to of of this library as it is
good for handling most of that would to and most of graphs that going to
important it to this kind of data for you may find that this library is going
to work want to things charts, graphs, histograms, bar graphs, plots, plots,
and
If you to your chart or graph to go through of data that you handling
during this working with matplotlib library is going to of options. it lack
of 3d that you may so this is to on your data. but for of basic parts that you
would to add into mix, and for most of visuals that you would to focus on,
you will find that matplotlib library is going to to it in no
 5:

What is

Now it is for us to look at basics that going to with and how can this for
our think about a of situations that you in past. for you shopping you
products that to did you it would for you, on product that you looking at,
or don of that you in past? if you similar to this on a that you shopping on,
you to
is you a call from a bank or financial company that will ask you to out a
policy or a loan. it is that this financial institution is going to calling this
would too long and would a going to to out who most to products starting.
So, a look at what is all about. is going to of parts of artificial that going to
focus mainly on from can on that has in past.

What do? it is going to and to that on data, than having do all of codings
to this programs or algorithms going to in a than what find in past can and
to to data. this is what is going to so strong and good at analyzing data that
you working with along way.
In world that work in today, or robots, to still to start following any of
instructions that you would but what if, in is to start on own from work
and as do. this is not that can right now, but it is an part that can to us of
our goals with on.

How works for data

is an option that is going to to work with data to us of information that this
option is going to a or work on and and that it can do all of this without on
how to in situation.

What this is that of writing out want, will data into a algorithm. from
algorithm is going to to build up logic on data that is a lot of parts that
with and figuring out how can work with this and can that most out of this
along way as
a lot of applications growing so much. if you with a to find familiar with
of work that can do with if you with kind of it is on your or of assistant
you will find that you with as and just start of what can do with as
In this is going to go into and us with products and that can’t today. it is to
do so many things for us, and think about of this as and jump on it and
how to it, to amazing things. with this in mind, to a bit about and how can
this for our as

What do?

algorithms that can with will with of data for training, to that would put in
input data to this algorithm, it is going to of as its basis for providing us
with of that along way.

that algorithm can for us is going to for accuracy. if find that accuracy is
this algorithm with is going to if that accuracy is not good, this algorithm
will go through again and until it accuracy that would
a lot of ways that is going to to us out. many analysis part of our data is
going to with and that can a in insights that going to find along way as and
with all of and that can as will that going to that

of
thing that to on is of that can work with. main of that can work with. going
to and going to to look at of and what can do with of to us our goals.

First, to a look at what is all about. talking about talking about can data
and to algorithm work. can this algorithm by a going to a of data that is
going to act as and of this or data is going to train or that working with. to
go through and train can it to start making or data is to it on.
can work with of as this is going to a bit than what to with but it is going
to add in of functionality that looking for it is to our work with of through
a lot of with of it can find of that in data.
is of data that want to work with, it is automatically to find of and that in
of data, usually through of in it. what it is not to do is add in to at all. this
that it is not to us that group is and is but it can a bit
say that had a bunch of that to bananas, and with this is going to a look at
and that found in and will dividing up of data into various if you find that
data is into it is all it is going to that data and add it into of that you

And finally, can work with of on of this, into it too much, going to find
that looks a lot what going to with that about but main is that is going to
on of trial and to algorithm to
As can a lot of parts that going to with that can work with. and this is
going to a important option that can focus on it to data if you looking to
work on a data of any kind, and you or algorithms that going to us to sort
through data and us insights and that you to working with that did

and how it fits with data

can an important part of data as long as it

as go through this that part of data is working on data analysis. this us to a
lot of data along way, and actually insights and that of it. to this to to our
(that can sort through all of data), find and us with our insights.

To and to that work way that want, to a of good algorithms in and this is is
going to into play a bit. you will find that with of and of algorithms that in
can that can go through any kind of data it is big or small, and us with that
is a that can to or working with think in a that humans do. this allows
algorithm to go through and find in that a human would to do, but it can
do it much and than any human could do manually.
Think about how hard this would to do manually for any human, or for a
group of who trying to through all of that data. it could to through all of
that data and find insights that and with how fast data is and and insights
would by got to that point anyway.
can this so much it allows us to a way to think through data and find and
insights that for our with right algorithm, can how works, and all of that to
this for us. with this in mind, it is to a look at and all of parts that to know
to this work for our

How work with data analysis?

Now that know a bit about how works and why it is important, it is for us
to a look at how can in and us out with our data analysis. so many why
can it to data analysis, so it is important to to look at how can it as
is going to for all of algorithms that want to along way. no how or how
your algorithm will a lot of coding and that with it going to run by that
will talk about in this and with of python, you can amazing algorithms that
us to sort through data.
So, if you hoping to go through this of data analysis to sort through your
data and what is found of it, you to a bit about of good with this is that is
going to to work with python that about that can it with a coding though
that with going to a bit
first of that to a look at is known as this is going to most basic form of that
can work with, but it will us with of parts that to things going and can us
to train our algorithms quickly and

To start, is simply going to of an algorithm to to map an input to a

particular output. going to or on this showing lots of with to algorithm in
that it will find and training is algorithm will to look at inputs, without
output, and us right on its own.
This is going to work on a data that was if mapping is algorithm is going
to to if it is not goals that to go through and to our algorithm to it
algorithms, will to good for data on in
This is going to a similar that would with a to is going to a who can to
from books and is going to and, if will pass. if not, will things up and will
to so that can from that in past so that this is going to basics that with
using

of that can work with is known as this is going to a that can in data
analysis it will to go through and classify both and without having to go
through and or with any information about of about
things or that our going to to classify going to such as purchasing of of of
of and things attacks or fraud with a bank. main that going to to with this
kind of is that want to our to of data that and allowing algorithm to to and
from data. to to to program how it can from that data.
It is common for a to to how to of of data, both and and what insights in it
may almost to prior information about all of data that a could a of and
working with this kind of can to things you not to train your of to it.

all of this in mind, will find that is not going to all that in such as working
with to a constant amount of information about data that is for hacking
attacks on a bank or a financial going to go through and and would with
up, but is going to to this.
In and is going to to go onto a and quickly from all of data from attack to
up. it can and insights about attacks, also actions to work with along way.
a lot of will want to work with any that you want to work with a program
or a that to do at a bit of on its own to things is going to right option to
focus on.

third of that to a look at is known as this is going to a bit to what saw with
two options, but a lot of can this kind of to us out with sorting through our
data, including our data analysis. into basics of and how can it for our

To start, in of artificial is going to a of dynamic programming that can

train algorithms, on of algorithm right and a of it wrong
of algorithms that or is going to to how to with that is going on around it.
or algorithm is going to kind of it but it it is going to kind of or in through
and is going to without any kind of from a human, by maximizing its and
figuring out way to that it is going to with.
algorithm is going to with this it has a to right way to and wrong way to it
through right way to it will with actions or to and it a for doing wrong, it
is going to this as and will how to avoid along way as
is going to of that can with and inspiration for it is going to found in
psychology. can this in a that is similar to how a child can a task. this is
going to contrasts to how options will approach a situation this particular
algorithm is not going to told how it should a task. it has to and go through
this all on its own.
As an which could a program that is up to play or a car, is going to with
that is all around it, and it is going to a of on how it can so, if can win of it
will a
This way as if not in that it should, that that it win it should or that is
wrong on programming, it is going to a of sort. in of it is going to than
winning, and it can from that in
through and is going to to good to its and of it a through dynamic
programming. of working with this kind of approach, work with artificial
is that it is going to allow our ai program to without programming having
to go through and out how an should its tasks.

As can a lot of parts that with of and to of and what can do with main of is
going to important on how want to this in our own data analysis. to look at
how can all of of for our own and from into using it to pick out algorithm
to
 Conclusion

Thank you for this book. is to by how data is going to to work for your
you will find that a lot of ways that you can amount of info that you to,
and all of data that you to data is just first to also to that can gain all of
insights and that out of that information, and this is of data is going to into
play.
This has to what data is all about and how it can your company in so many
ways. at of tasks that data can out with, what data is and how to work with
of data, of data, and so much this us to of parts that to data analysis, and
how and using all of that information can to grow your
But this is not only that can work with. also to this a bit and not just data,
but also to that data and what information it holds. this is a part of data but
it without it, data would just sit without

Data is a thing to add to your and it can you to satisfaction is high, that is
low, that you can and can with in such as what products you should and
put out on but all of this is not that just on its own. working with data and
adding in python and that with it can you to work with python data to
many of your own and to out to out this to you with it right away.
you to about what data analysis can do for you, and how you can work
with this data analysis to along with python and it is to a look at what this
has to
and tools will for data analysis, and of will not analyst.
staffing for data analysis will on and from to analysts to to in of data will a
crunch in availability of big data might tactics. that various organizations
will training to a having big data in form of can on horizon.
Many jumping on board and about how to work with data and what all
this can do for and with of this you will to do of as you to jump on board
and about data and how you and your can from this to out this to
Data is taking world, and many no what industry in, found that this kind
of is what to not only data but also to it and an analysis to find insights
and that data is and add in python to and that is going to to find way to
that our standing in industry.
a lot of parts that with data and to put all can us to do with our finding
products to bring to and and with of this can find ways to out and that will
work for us. it and a good data analysis with right algorithms from python,
but it can of ways to smart and sound for your

Many going to of data to as many as company will to this and will find
that it can add to bottom profits, and in and to about how data works, and
all of parts that with it, to out this to