Python Pandas

www.sunilos.com
www.raystec.com
Pandas Library
Lets play with Tabular Data
6/1/2020 www.SunilOS.com 1

What is Pandas ?
 Pandas is open source.
 BSD- licensed Python library providing high – Performance.
 Easy to use for data structures and data analysis.
 Pandas use for different types of data.
o Tabular data with heterogeneously-typed columns.
o Ordered and unordered time series data.
o Arbitrary matrix data with row & column labels
o Unlabeled data
o Any other form of observational or statistical data sets

Pandas is Data Frame
max std filter min count

Features of pandas
 It has a DataFrame Object with customized indexing.
 We can reshape data sets.
 We can perform aggregations and Transformations on the data sets
 It is used for data alignment and integration of the missing data.
 It provides the functionality of Time Series.
 Process a variety of data sets in different formats like matrix data,
tabular heterogeneous, time series.
 Handle multiple operations of the data sets such as subsetting,
slicing, filtering, groupBy, re-ordering, and re-shaping.
 It integrates with the other libraries such as SciPy, and scikit-learn.
 Provides fast performance, and If you want to speed it, even more,
you can use the Cython.

Usage of Pandas
 Modeling
 Visualization
 High Performance Computing
 Big Data
 Statistical Computing
 Numerical Computing
 Data Mining
 Text Processing
 Computing Language
 Educational Outreach
 Computational thinking

How to install pandas
 You can run following command to install pandas.
o pip install pandas
o OR
o conda install pandas
 After installation you can check version of pandas by running
the following command at python script mode.
o Import pandas
o pandas.__version__

Data Structures
There are 2 data structures in pandas.
o Series.(One D array)
o DataFrame

Series Data Structure
Series is a one –dimensional array.
It contain homogeneous data.
Size is fixed.
10 20 30 40 50 60 70

Series Data Structure Cont.
Series is One Dimensional array to hold different types of
values.
o pandas.Series( data, index, dtype, copy)
data: It can be any list, dictionary, or scalar value.
index: The value of the index should be unique and
hashable. It must be of the same length as data. If we do not
pass any index, default np.arrange(n) will be used.
dtype: It refers to the data type of series.
copy: It is used for copying the data.

Creating a Series
We can create Series by using various inputs:
o Array
o Dictonary
o Scalar value

Using Array
import pandas as pd
import numpy as np
info =
np.array(['R','a','y','s','T','e','c',
'h'])
a = pd.Series(info)
print(a)

Using Dictionary and scalar Value
Using dictionary
o import pandas as pd
o info = {1:'Rays',2:'Tech'}
o a = pd.Series(info)
o print(a)
Using Scalar Value
a = pd.Series(4,index=[1,2,3,4,5])
print(a)

Series Object attribute:
 import pandas as pd
x = pd.Series([1,2,3],index = ['a','b','c'])
 # It returns a tuple of shape of the data.
print("Shape:",x.shape)
 # It returns the size of the data.
print("Size:",x.size)
 # Defines the index of the Series.
print("Index:",x.index)
 # It returns the number of dimensions in the data.
print("Dimension:",x.ndim)
 # It returns the values of series as list
print("values:",x.values)


Series Object attribute (cont.)
 # It returns the data type of the data.
print("Data Type:",x.dtype)
 # It returns the number of bytes in the data
print("Bytes:",x.nbytes)
 # It returns True if Series object is empty, otherwise
returns false.
print("is empty:",x.empty)
 #It returns True if there are any NaN values, otherwise
returns false.
print("is empty:",x.hasnans)

Pandas Operations
map()
o import numpy as np
o a = pd.Series(['Java', 'C', 'C++', np.nan])
o print(a.map({'Java':'Core',"C":'Python'}))
std()
o print("Series:",np.std([1,2,3,4,5]))
count()
o i = pd.Series([2, 1, 1, np.nan, 3,4,5,5])
o print(i.count())

Pandas Operations
filter()
o arr=np.array(([20,30,40],[50,70,80]))
o df=pd.DataFrame(arr,index=["Vijay","Jay"],columns=
["Physics","Chemistry","Maths"])
o print(df)
o print(df.filter(items=["Physics"]))

Data Frame & Panel Data Structure
 DataFrame is a two–dimensional array.
 It contain heterogeneous data.
 Size is fixed.
 Data Type of Columns.
 Panel is three-dimensional
array.
 It is hard to represent the Panel
In graphical representation.
 But panel is consider as a Data
Frame.
NAME AGE ADDRESS
Rahul 22 Indore
Ram 35 Ayodhya
Krishna 25 Mathura
COLUMN NAME TYPE
NAME String
AGE Integer
ADDRESS String

Creating DataFrame
 # importing the pandas library
import pandas as pd
#empty DataFrame
df = pd.DataFrame()
print (df)
# Using List
list=["Rays","Tech","SunilOs"]
print(pd.DataFrame(list))
# using Dict
info = {'ID' :[101, 102, 103],'Department'
:['B.Sc','B.Tech','M.Tech',]}
print(pd.DataFrame(info))

Creating DataFrame
#Using Dict of series
info = {'one' : pd.Series([1, 2, 3, 4, 5, 6],
index=['a', 'b', 'c', 'd', 'e', 'f']),
'two' : pd.Series([1, 2, 3, 4, 5, 6, 7, 8],
index=['a', 'b', 'c', 'd', 'e', 'f', 'g',
'h'])}
d1 = pd.DataFrame(info)
print(d1)
one two
a 1.0 1
b 2.0 2
c 3.0 3
d 4.0 4
e 5.0 5
f 6.0 6
g NaN 7
h NaN 8

DataFrame Implementation
import pandas as pd
a list of strings
x = ['Rays','Tech',"SunilOs"]
Calling DataFrame constructor on list
df = pd.DataFrame(x)
print(df)

Column Selection from a DataFrame:
 from pandas import DataFrame,Series
 x={'Id':Series([1,2,3,4,5]),
 "Name":Series(["A","B","C","D","E"])}
 df=DataFrame(x)
 print(df)
 print(df['Id'])
 print(df["Name"])

Column Addition and Deletion to DataFrame
x={'Id':Series([1,2,3,4,5]),"Name":Series(["A"
,"B","C","D","E"])}
df=DataFrame(x)
print(df)
 # ADD a Column from DataFrame
df['Subject']=Series(["C","C++","Java","python
","Java","Julia"])
 del df["Id"] # Delete a Column from DataFrame
Using del function
print(df)
 df.pop("Name") # Delete a Column from
DataFrame Using pop function

Row Addition from DataFrame
 x={'Id':Series([1,2,3,4,5],index=['a','b','c','d
','e']),
"Name":Series(["A","B","C"],index=['a','b','c'])
}
 df=DataFrame(x)
 d=DataFrame({"Subject":Series([1,2,3],index=['a'
,'b','c'])})
 print(df.append(d))
#OR
 d=DataFrame([[1,"Ajay","Python"]],
columns=["Id","Name","Subject"])
 print(df.append(d))

Row Deletion from DataFrame:
from pandas import DataFrame,Series
x={'Id':Series([1,2,3,4,5],index=['a','b
','c','d','e']),
"Name":Series(["A","B","C"],index=['a','
b','c'])}
df=DataFrame(x)
print(df.drop('a'))

Row Selection from DataFrame
 x={'Id':Series([1,2,3,4,5],index=['a','b','c
','d','e']),
"Name":Series(["A","B","C"],index=['a','b','
c'])}
 df=DataFrame(x)
#by row label
 print(df.loc['a'])
#by using row integer row location
 print(df.iloc[4])
#by using slice operator :
print(df[0:3])

DataFrame Object Methods:
append():Add the rows of other dataframe to the end of the
given dataframe.
apply():Allows the user to pass a function and apply it to
every single value of the Pandas series.
count(): Count the number of non-NA cells for each column
or row.
describe():Calculate some statistical data like percentile,
mean and std of the numerical values of the Series or
DataFrame.

DataFrame Object Methods: (cont.)
 drop_duplicate(): Remove duplicate values from the DataFrame.
 head():Returns the first n rows for the object based on position.
 mean(): Return the mean of the values for the requested axis.
 merge():Merge the two datasets together into one.
 pivot_table():Aggregate data with calculations such as Sum,
Count, Average, Max, and Min.
 query():Filter the dataframe.
 Sort_index(): Sort the dataframe.
 sum(): Return the sum of the values for the requested axis by the
user.
 to_excel(): Export the dataframe to the excel file.
 transpose(): Transpose the index and columns of the dataframe.

Importing Data:
pd.read_csv(filename) : It read the data from CSV file.
pd.read_table(filename) : It is used to read the data from
delimited text file.
pd.read_excel(filename) : It read the data from an Excel file.
pd.read_sql(query,connection _object) : It read the data from
a SQL table/database.
pd.read_json(json _string) : It read the data from a JSON
formatted string, URL or file.

Time-Series in Panda
 Time-series data is a data with equal no of intervals such as daily,
monthly, yearly.
 Data is collected on a fraction of time
o from datetime import datetime
o range_date = pd.date_range(start ='1/1/2019', end
='1/08/2019', freq ='Min')
o print(range_date)
o print(len(range_date))

Panda Date Time
 import pandas as pd
 dmy = pd.date_range('2017-06-
04', periods=5, freq='S')
 We can Format the String to date
o import datetime
o print(pd.to_datetime('18000706', format='%Y%m%d',
errors='ignore') )
o print(datetime.datetime(1800, 7, 6, 0, 0) )
o print(pd.to_datetime('18000706', format='%Y%m%d',
errors='coerce') )


Pandas time offset
It performs various operation on time i.e.
adding and subtracting
o # Create the Timestamp
o p = pd.Timestamp('2018-12-12 06:25:18')
o # Create the DateOffset
o do = pd.tseries.offsets.DateOffset(n = 2)
o # Add the dateoffset to given timestamp
o new_timestamp = p + do
o # Print updated timestamp
o print(new_timestamp)

Pandas Time Period
It represents the time period i.e. day, month, quarter,
year.
It converts frequency to time period.
o daily=pd.period_range("2000","2010",freq='
D')
o print(daily)

www.sunilos.com
www.raystec.com
Image Processing

Image Processing
Image Processing is a process to perform some
operation on the image such as:
o Change color of the image
o Resize image
o Crop the image etc.
There are many libraries available for image
processing. Some common libraries are
following: we will Use OpenCv for Image
Processing.

Image Processing Libraries
 Scikit-image:
o It uses a Numpy array as an image object. Skimage transforms the image
file into n-dimensional array objects. The ndarray can be type of integer
and float. It is fast because written in c
 OpenCV:
o It is written in C++ and Python. Opencv can work with Numpy, scipy and
matplotlib. It is used in Image processing, face detection, Object detection
etc.
 Mahotas:
o It uses advanced features such as image processing. It can process 2D and
3D images. It can extract information from images. It has more than 100
functions for computer vision processing.

Image Processing Libraries
 SimpleiTk:
o It treats images as set of points in a space. It support many dimensions such
as 2D, 3D and 4D
 SciPy:
o Scipy has a module named scipy.ndimage for image processing. Scipy
basically used for scientific calculations
 Pillow:
o It support many image formats such. It is advanced version of PIL.
 Matplotlib:
o It is used for visualization or graph plotting, but we can use it for image
processing. It does not support all image file formats. It can used for image
altering

What is OpenCV?
OpenCV is a Python open-source library, which
is used for computer vision in Artificial
intelligence, Machine Learning, face recognition,
etc.


How OpenCV Works:
Human eye provides information whatever he
sees. But the machine knows only numbers. The
Machine converts images into numbers on the
basis of pixel data and stores them in memory.
The picture intensity represented as a number.

How OpenCV Works:

Representation of Image
 Gray scale: black and white images
 RGB: Colorful images

Why is OpenCV used for Computer Vision?
Free of cost
It is fast.
Requires less.
It is portable.

How to install OpenCv:
Open command prompt and type
o pip install opencv-python
Open conda prompt
o conda install -c conda-forge opencv in
anaconda prompt


Supported File Format
Window bitmaps - *.bmp, *.dib
JPEG files - *.jpeg, *.jpg, *.jpe
Portable Network Graphics - *.png
Portable image format- *.pbm, *.pgm, *.ppm
TIFF files - *.tiff, *.tif

Reading and writing Images
 import cv2
 # read a file as grayscale
 img=cv2.imread("Desert.jpg",0)
 cv2.imshow("Image",img)
 cv2.waitKey(0)
 # write to a file
 res=cv2.imwrite("copy.jpg",img)
 print("status:",res)

OpenCV Basic Operation on Images:
Access pixel values and modify them
Access Image Properties
Setting Region of Image
Splitting and merging images
Change the image color

Accessing Image Properties
import cv2
img = cv2.imread("Desert.jpg",1)
print("Height:",img.shape[0])
print("Width:",img.shape[1])
print("channel:",img.shape[2])
print("size:",img.size)

Making Borders for Images:
 import cv2 as cv
 BLUE = [255,0,0]
 img1 = cv.imread("Desert.jpg",1)
 img =
 cv.copyMakeBorder(img1,50,10,10,10,cv.BORDER
_CONSTANT, value=BLUE)

OpenCV Resize the image:
 import cv2
 img = cv2.imread("Desert.jpg", 1)
 scale = 60
 width = int(img.shape[1] * scale / 100)
 height = int(img.shape[0] * scale / 100)
 dim = (width, height)
 print("Initial dimension ",img.shape)
 # resize image
 resized = cv2.resize(img, dim)
 print('Resized Dimensions : ', resized.shape)
 cv2.imshow("Resized image", resized)
 cv2.waitKey(0)
 cv2.destroyAllWindows()

OpenCV Drawing Functions
 cv2.circle():
o cv2.circle(img, center, radius, color[,thickness [, lineType[,shift]]])
 cv2.rectangle():
o cv2.rectangle(img, pt1, pt2, color[, thickness[,lineType[,shift]]])
 cv2.ellipse():
o cv2.ellipse(img, center, axes, angle, startAngle, endAngle, color[, thickness[,
lineType[, shift]]])
 cv2.ellipse:
o Cv2.ellipse(img, box, color[, thickness[, lineType]])
 cv2.line:
o cv2.line(img, pt1, pt2, color[, thickness[,lineType[, shift]]])
 Put text:
o cv2.putText(img, text, org, font, color)
 To Draw Polylines:
o cv2.polyLine(img, polys, is_closed, color, thickness=1, lineType=8, shift=0)

Drawing some shapes on image
 import numpy as np
 import cv2
 img = cv2.imread("Desert.jpg",1)
 cv2.rectangle(img,(15,25),(200,150),(0,255,5
5),15)
 cv2.imshow('image',img)
 cv2.waitKey(0)

Write Text to Image:
 import cv2
 font = cv2.FONT_HERSHEY_SIMPLEX
 # Create a black image.
 img = cv2.imread("Desert.jpg",1)
 cv2.putText(img,'Rays
Technology',(10,50),font,2,(100,255,255),2)
 #Display the image
 cv2.imshow("image",img)
 cv2.waitKey(0)

Crop the Image
import cv2
img=cv2.imread("Desert.jpg")
y=300
x=200
h=100
w=200
crop = img[y:y+h, x:x+w]
# crop=img[300,500]
cv2.imshow("Image:",crop)
cv2.waitKey(0)

Crop rectangle image to circle
 import cv2
 img=cv2.imread(“Koala.jpg")
 img1=cv2.resize(img,(300,300))
 h,w,c=img1.shape
 print(h,w,c)
 center=(w//2,h//2)
 radius=180
 color=(255,255,255)#white
 thickness=80
 cv2.circle(img1,center,radius,color,thickness)
 cv2.imshow("Cropped",img1)
 cv2.waitKey(0)
 cv2.imwrite("Koala1.jpg",img1)

OpenCV Resize the image:
 import cv2
 img = cv2.imread("Desert.jpg", 1)
 scale = 60
 width = int(img.shape[1] * scale / 100)
 height = int(img.shape[0] * scale / 100)
 dim = (width, height)
 print("Initial dimension ",img.shape)
 # resize image
 resized = cv2.resize(img, dim)
 print('Resized Dimensions : ', resized.shape)
 cv2.imshow("Resized image", resized)
 cv2.waitKey(0)

Making Borders for Images:
 import cv2 as cv
 BLUE = [255,0,0]
 img1 = cv.imread("Desert.jpg",1)
 replicate
=cv.copyMakeBorder(img1,50,10,10,10,cv.BORDE
R_CONSTANT, value=BLUE)
 cv.imshow("replicate",replicate)
 Cv.waitKey(0)

OpenCV Canny Edge Detection:
 import cv2
 img = cv2.imread('Koala.jpg')
 edges = cv2.Canny(img, 100, 200)
 cv2.imshow("Edge Detected Image", edges)
 cv2.imshow("Original Image", img)
 cv2.waitKey(0) # waits until a key is pressed

www.sunilos.com
www.raystec.com
Python Numpy
The fundamental package for scientific computing with Python

What is Numpy
Numpy stands for numeric python.
Numpy module is used to perform computation
and processing on multidimensional and single
dimensional arrays.

Advantages Of NumPy
It is used for array Oriented Computing.
High efficiency in a multidimensional array.
Use for scientific computation.
We can reshape the data which is stored in
multidimensional array
NumPy provides the in-built functions for linear
algebra and random number generation.

NumPy Environment Setup
NumPy doesn't come bundled with Python. We
have to install it using the python pip installer.
Execute the following command.
o pip install numpy
OR
The Anaconda provides SciPy stack which is a
free distribution of the Python SciPy package

Numpy Ndarray
Ndarray is an object to
store similar kinds of
elements.
We can create a
collection of elements
of the same data type.
The array index starts
from 0.

How to Create Ndarray
 To create ndarray first import the numpy module and use
array() function.
o a=np.array
o print(a)
o za=n.zeros((2,3),dtype=int)
o print("Zeros array:n",za)
o oa=n.ones(shape=(3,2))
o print("Ones arrayn",oa)
o emp=n.empty(shape=(1,2),dtype=int)
o print("empty array:n",emp)

Array From existing data:
 numpy.array(object, dtype = None, copy = True,
order = None, subok = False, ndmin = 0)
 object- it can be set, tuple dictionary etc.
 dtype-it represents the type of the object. By default the
type is none
 copy- It means the object is copied. By default value is
true. It is an optional attribute.
 order- we can arrange arrays in 3 possible ways C(column
order), R(row order), A(any).
 subok- by default the base class is array, by making this
attribute true we can make subclasses
 ndim- To define the minimum dimension of the array.

Array From list data:
 ar=np.array({1,2,3})#OR
 ar=np.array([1,2,3],float)#OR
 ar=np.array((1,2,3),complex)
 l=[[5,6,7],[1,2,3],[8,9,7]]
 la=n.asarray(l)
 print("Array from List:n",la)
 print(ar)
 print(type(ar))

Array from tuple
t=((2,2,3),(1,2,3))
a=n.asarray(t,dtype=float)
print("Array From tuple:n",a)

Ndarray Object attributes:
ndim-ndim function is used to find the dimension of the
given array.
dtype- is used to get the type of the elements of the
array
shape- It is used to get the shape of the given array.
Menas no of rows and columns in a array.
size- is used to get the no of elements of the array of size
of the array
Itemsize- to find the size of each element in the array.

Array Attribute
a=np.array([[1,2,3],[2,3,4],[4,5,6]])
print(a)
print("Dimension:",a.ndim)
print("Data type:",a.dtype)
print("Size of array:",a.size)
print("Shape of Array:",a.shape)
print("Size of elements:",a.itemsize)

Restructure the shape of the ndarray:
a=np.array([[1,2],[2,3],[5,6]])
print("Original array")
print(a.shape)
ra=a.reshape(2,3)
print(ra)

Slicing in the Array-
To extract the range of elements from a array we
can use slicing. It is same as we have used in the
list
a=np.array([1,2,3,4,5,6])
print(a[1:3])

Numpy array using numerical Range:
Arange: The arange() function is used to create
an array of evenly spaced over the given range.
Linspace: The linspace() function is used to get
evenly spaced elements over the given interval
range.
Logspace: lgapace() function is used to create an
array of evenly spaced values according to log
scale.


Numpy array using numerical Range:
a=np.linspace(1,20,5)
print(a)
a=n.arange(0,10,2,float)
print("Ranged Array:n",a)
l=n.logspace(2,10,5)
print("LogSpaced Array:n",l)

Aggregate Operations on array
We can perform aggregate function on the array
such as: sum(), max(), min() etc.
from numpy import *
a=array([1,2,3,4,5])
print(type(a))
print("Sum:",sum(a))
print("Max:",max(a))
print("Min:",min(a))
print("Std:",std(a))
print("Square Root:",sqrt(a))

Arithmetic Operations on array
To perform arithmetic operations the shape of
the array must be same
 from numpy import *
 a=array([[1,2,3],[2,3,4]])
 b=array([[2,4,6],[4,6,8]])
 print("Addition:",a+b)
 print("Subtraction:",a-b)
 print("Multiplication:",a*b)
 print("Division:",a/b)

Ndarray Axis
 We can perform operations row and column wise on array.
 from numpy import *
a=array([[1,2,3],[1,3,4]])
print(a)
print("Max:",a.max(axis=1))
print("Max:",a.min(axis=1))
print("Max:",a.sum(axis=1))

Array concatenation
we can concat multidimensional arrays vertically
and horizontally
from numpy import *
a=array([[1,2,3],[1,3,4]])
b=array([[5,6,7],[8,9,10]])
print("Vertically
concat:n",vstack((a,b)))
print("Horizontally
concat:n",hstack((a,b)))

Iteration over the Array
nditer() function is used to iterate through
arrays.
import numpy as n
na=n.array([[1,2,3],[5,6,7]])
print("Array:n",na)
for x in n.nditer(na):
 print(x,end=" ")

Numpy String Function
Sr. no Function Description
1 add() It is used to add string type element to array
2 multiply() It is used to obtain the multiplied copy of the string element
3 center() It is used to make a center aligned string filled with specified no
character.
4 capitalize() It returns the copy of the string with the first letter capitalized.
5 title() It returns the title case string.
6 lower() It returns a string with lowercase letters .
7 upper() It returns a string with uppercase letters.
8 split() It returns the words from string.
9 splitlines() It breaks the list of lines
10 strip() It removes the left and right spaces.
11 join() It joins the given sequence of strings.
12 replace() It replaces all the occurring words of substring in a string.
13 decode() It is used to decode the string in specified format.
14 encode() It is used to encode a string into a specified codec.

Numpy String Function(cont.)
 str=["Vijay","Ajay","Sanjay"]
 str1=[" Chouhan"," Sharma"," singh"]
 print("Original array:n",str)
 print("Add:n",np.char.add(str,str1))
 print("Multiply:n",np.char.multiply(str,3))
 print("Center:n",np.char.center(str,6))
 print("Strip:n",np.char.strip(str))
 enstr=np.char.encode("Welcome","cp500")
 print("Encode:",enstr)
 print("Decode:",np.char.decode(enstr,"cp500"
))

Numpy Mathematical Function:
Sr. No Function Description
1 sin() Used to calculate sin of given angles
2 cos() Used to calculate cosine of given angles
3 tan() Used to calculate tangent of given angles
4 arcsin() Used to calculate inverse of sin of given angles
5 arccos() Used to calculate inverse of cosine of given angles
6 arctan() Used to calculate inverse of tangent of given angles
7 degree() It is used to convert the angles in degree
8 around() It is used to round off the value according to the given number.
9 ceil() It is used to calculate the ceil value of the given no
10 floor() It is used to calculate the floor value of the given no.

Numpy Function:
Sr. No Function Description
1. array() To create a n dimensional array
2. append() To add new values to existing array
3. concatenate() To concatenate arrays into a single array
4. reshape() It is used to change the shape of the existing array
5. sum() It is used to get the sum of the whole array or row
and column wise sum of arrays.
6. random() It is used to get the array of random no according to
the called function.
7 zeros() It returns the array containing zeros of given shape.
8. log() It is used to calculate log of each elements in the
array
9. transpose() It is used to change the shape of the array
10 unique() It is used to get the unique elements from the array.

Numpy Function:
11. std() It is used to calculate the standard deviation along
the axis.
12 mean() It is used to calculate arithmetic mean to given
axis
13. sort() It is used to sort the element of the given array.
14 average() It is used to calculate the weighted average of the
array
15 arange() It is used to get an array according to given start,
stop and steps
16 linspace() It is used to get the evenly spaced array according
to given size.
17 trunc() It is used to truncate the value of the given array
18. asarray() It is used to convert the python sequence into
numpy array

Numpy Data Types-
numpy provides the wide range of datatypes. We
can create the dtype object as following:
numpy.dtype(object, align , copy)
o d=np.dtype([('salary',np.float)])
o arr=np.array([(10000.12,),(20000.50,)],dtype=d)
o print(arr['salary'])

www.sunilos.com
www.raystec.com
Matplotlib

Matplotlib
It is a python library developed by John D. Hunter
in 2002.
 Latest release is 3.1.1 released in 2019.
It is used for data visualization on multi platforms
It can be used in python script, shell, web
application, and other python gui toolkits.

Matplotlib Architecture
 There are three different layers in the architecture of the
matplotlib which are the following:
 Backend Layer- This is the bottom layer which is responsible
for implementation of the function which is required for
plotting. There are 3 important classes- FigureCanvas(surface
to draw figure) , Renderer(responsible for drawing on surface),
Event(handles mouse and key events).
 Artist layer - This is the middle layer of the architecture. It
takes care of all the plotting function
 Scripting layer- This is the top most layer where all the code
runs.

Different part of the Matplotlib:

Different part of the Matplotlib:
Figure: It is a canvas that holds plots.
Axes: a figure can contain several plots.
Axes is a plot.
Axis: axis represents the limit of the plots..
Artist: artist are all visual objects which are
shown on the figure such as text objects,
lines and collection.

Environment setup:
If you are installed in anaconda distribution then
matplotlib is pre-installed in the anaconda distribution
no further installation required. Or we can install using
pip command as following:
o Pip install matplotlib
To know installed version:
o import matplotlib
o matplotlib.__version__

Types of Graphs:
We can plot different kinds of graphs in
matplotlib with different styles. Some of the
graph as following:
Bar graph
Scatter graph
Line Graph
Pie Charts
Histograms
Box plot

Some Important Functions in Matplotlib:
Sr. no. Function Description
1 axes() T is used to add axes(plot) to a given figure.
2 bar(values, color) It is used to create the horizontal bar graph with
given values and color
3 barh(values, color) It is used to create horizontal bar graphs with given
color and values.
4 figure() It is used to create a figure class instance, and
handle figure level attribute
5 grid(bool) It is used to show the grids on the plots
6 hist(values, bin) It is used to create a histogram
7 legend() It is used to show the legend on the axis which
describes the elements of the graph
8 pie(values,categorical values) It is used to create a pie chart.
9 plot(x values, y values) It is used to plot the simple line graph with given x
and y axis values
10 plot3D(x,y,z) It is used to draw a 3D graph with given axis
values.

Some Important Functions in Matplotlib:
11 subplot(row,col,index) It is used to create a subplot for a given figure.
12 suptitle(str) It is used to set the common title for all the plots
13 scatter(x values, y values) It is used to create a scatter plot.
14 set_xlabel(str) This is the axes level method.It is used to give the
label to the x axis.
15 set_ylabel(str) This is the axes level method.It is used to give the
label to the y axis.
16 scatter3D(x values, y, value, z
values)
It is used to draw a scatter graph in 3 dimension
space
17 savefig(loc) It is used to save the plotted graph to the specified
Location
18 show() It is used to makeplotted graph visible
19 title(str) It is used to set the title of the plot
20 text() It is used to set the text to the plot
21 xlabel(str) It is used to set the label of the x axis of a plot
22 ylabel(str) It is used to set the label text of a y axis of a plot

Simple graph with matplotlib
import matplotlib.pyplot as plt
#plotting our canvas
plt.plot([1,2,3],[4,5,1])
#display the graph
plt.show()

Add title and label to the axis
import matplotlib.pyplot as plt
x = [3, 4, 8]
y = [5, 12, 2]
plt.plot(x, y)
plt.title('Line graph')
plt.ylabel('Y axis')
plt.xlabel('X axis')
plt.show()

Changing Color and style of the line in a graph
Matplotlib support following abbreviations for color: b
(blue), r (Red), g (Green), c (Cyan), m (Magenta),
y(Yellow),k (Black), w(white)
‘b’- to get a blue solid line with default shape.
‘ro’- Red color circle
‘-g’ - Green solid line
‘--’ - dashed line with default color.
‘^k:’ - Black triangle connected by dotted lines.
‘d:’ - Diamond shapes with default color, connected by
dashed line

 import matplotlib.pyplot as plt
 plt.title("Line Graph")
 plt.ylabel("Y axis")
 plt.xlabel("X axis")
 x=[1,2,3,4,5]
 y=[2,4,6,8,10]
 #data points with red color and square Shapes
 plt.plot(x,y,"rs")
 plt.show()

Line Graph
plt.title("Line Graph")
plt.plot([1,4,5],[5,3,2],'r',linewidth=5,
label="Line1")
plt.plot([1,5,3],[6,4,9],'k',linewidth=5,
label="Line2")
plt.legend()
plt.grid(True)
plt.show()

fill_between():
import numpy as np
x = np.arange(0.0, 2, 0.01)
# x = np.arange(1, 5, 1)
y1 = np.sin(2 * np.pi * x)
y2 = 1.2 * np.sin(4 * np.pi * x)
fig, ax = plt.subplots(1, sharex=True)
ax.plot(x, y1, x, y2, color='black')
ax.fill_between(x, y1, y2, where=y2 >= y1,
facecolor='blue')
ax.fill_between(x, y1, y2, where=y2 <= y1,
facecolor='red')
ax.set_title('fill between where')
plt.show()

Output

Plotting More Than one Graph in a single Figure:
Matplotlib provides the function to plot more
than one graph in a single figure.
subplot()
subplots()
subplot2grid()

subplot() Example:
 x=["Vijay","Jay","Ajay"]
 y=[90,80,70]
 plt.subplot(131)
 plt.bar(x,y)
 plt.scatter(x,y)
 plt.plot(x,y,"^k")
 plt.show()

subplot

subplots():
 this function provides a grid like structure to plot the subplots:
o plt.subplots(nrow,ncols)
o from matplotlib import pyplot as plt
o fig,a=plt.subplots(2,2)
o x=[1,4,3]
o y=[5,6,7]
o a[0][0].set_title("Line")
o a[0][0].plot(x,y)
o a[0][1].set_title("Sin")
o a[0][1].plot(x,np.sin(x))
o a[1][0].set_title("Cosine")
o a[1][0].plot(x,np.cos(x))
o a[1][1].set_title("tangent")
o a[1][1].plot(x,np.tan(x))
o plt.savefig("subplots.png")
o plt.show()

subplots

subplot2grid():
 with the help of this function we can vary the size of the plots.
o Plt.subplot2grid(shape, location, rowspan, colspan)

subplot2grid() implementation
a1 = plt.subplot2grid((3,3),(0,0),colspan = 2)
a2 = plt.subplot2grid((3,3),(0,2), rowspan = 3)
a3 = plt.subplot2grid((3,3),(1,0), colspan = 2)
a4 = plt.subplot2grid((3,3),(2,0),colspan = 2)
import numpy as np
x = np.arange(1,10)
a2.plot(x, x*x)
a2.set_title('square')
a1.plot(x, np.exp(x))
a1.set_title('exp')
a3.plot(x, np.log(x))
a3.set_title('log')
a4.plot(x, np.sin(x))
a4.set_title('sign')
plt.tight_layout()
plt.show()

Bar Graph:
 Bar graphs are mostly used to show association among the data.
The matplotlib provides a bar() function to draw a bar graph. We
have to pass the categorical variable , values and color.
 from matplotlib import pyplot as plt
x=["Ram","Shyam","Ghanshyam"]
y=[30,90,70]
plt.title("Bar Graph")
plt.xlabel("Name")
plt.ylabel("Marks")
plt.grid(True)
plt.bar(x,y,color="cyan")
plt.show()

Horizontal Bar Graph
 barh(): this function is used to draw a horizontal graph:
x=["Ram","Shyam","Ghanshyam"]
y=[30,90,70]
plt.title("Bar Graph")
plt.xlabel("Marks")
plt.ylabel("Name")
plt.grid(True)
# plt.bar(x,y,color="cyan")
plt.barh(x,y,color="Black")
plt.show()

Pie Chart:
 Pie chart is a circular graph. It is used to show the percentage of data
with the help of slices or segments of the circle.
x=["Jay","Vijay","Sanjay"]
y=[12,27,60]
#distance from center
ex=[0.1,0,0]
colors=["Gold","Silver","cyan"]
plt.pie(y,ex,labels=x,colors=colors,autopct='%1.
1f%%',shadow=True,startangle=90)
plt.title("Pie Chart")
plt.legend(loc="lower left")
plt.show()

Pie Chart (cont.)

Scatter Plot:
 It is Used to compare variables. The data is displayed as a points.
x = [5,7,10]
y = [18,10,6]
x2 = [6,9,11]
y2 = [7,14,17]
plt.scatter(x, y)
plt.scatter(x2, y2, color='g')
plt.title('Info')
plt.ylabel('Y axis')
plt.xlabel('X axis')
plt.grid(True)
plt.show()

Histogram Charts:
 There is a difference between the histogram and bar graph
 Histogram is Used to show distribution of data, bar graph is used
to show comparison of different entities.
 A histogram is a type of bar plot that shows the frequency of a
number of values compared to a set of values ranges.
 There is four histtype available:
o bar-by default
o barstaked- multiple data stacked on top of each other
o step- line plot by default unfilled
o stepfilled- line plot with filled area.


Histogram implemenatation
# from matplotlib import pyplot as plt
Marks =
[21,53,60,49,25,27,30,42,40,1,2,102,95,8,15,
105,70,65,55,70,75,60,52,44,43,42,45]
bins = [0,10,20,30,40,50,60,70,80,90,100]
plt.hist(Marks, bins, histtype='bar',
rwidth=0.8)
plt.xlabel('age groups')
plt.ylabel('Number of people')
plt.title('Histogram')
plt.show()

3D graph plot:
Firstly the matplotlib package is developed to
draw 2D graphs.
But after 1.0 version to draw 3D graphs all the
utilities are added above the 2D sets.
Three-dimension plots can be created by
importing the mplot3d toolkit, include with the
main Matplotlib installation:
o from mpl_toolkits import mplot3d

3D graph (cont.)
 from mpl_toolkits import mplot3d
import numpy as np
import matplotlib.pyplot as plt
fig = plt.figure()
ax = plt.axes(projection='3d')
x=np.arange(10)
y=np.sin(x)
z=np.cos(x)
ax.plot(x,y,z)
ax.set_xlabel("X axis")
ax.set_ylabel("Y axis")
ax.set_zlabel("Z axis")
print(x)
plt.show()

Output

Working with text:
We can use a subset TeXmarkup in any
Matplotlib text string by placing it inside a pair
of dollar signs ($).
To make subscripts and superscripts, use the '_'
and '^' symbols −


Working with text (cont.)
 plt.title("How to add Text ")
 plt.text(0.29,0.18,"$E=MC^2$",
color=“blue", fontsize=50,
style="italic",
bbox = {'facecolor': 'red'})
 plt.show()

Disclaimer
This is an educational presentation to enhance the
skill of computer science students.
This presentation is available for free to computer
science students.
Some internet images from different URLs are used
in this presentation to simplify technical examples
and correlate examples with the real world.
We are grateful to owners of these URLs and
pictures.

Thank You!
www.SunilOS.com

Python Pandas

Recomendados

Recomendados

Mais conteúdo relacionado

Mais procurados

Mais procurados (20)

Semelhante a Python Pandas

Semelhante a Python Pandas (20)

Mais de Sunil OS

Mais de Sunil OS (20)

Último

Último (20)

Python Pandas

Notas do Editor