Difference between revisions of "Python for Data Science"

From Sinfronteras
Jump to: navigation, search
(→‎NumPy)
(→‎NumPy)
Line 81: Line 81:
 
<br />
 
<br />
 
===NumPy===
 
===NumPy===
−
* NumPy is a Linear Algebra Library for Python, the reason it is so important for Data Science with Python is that almost all of the libraries in the PyData Ecosystem rely on NumPy as one of their main building blocks.
+
* NumPy (or Numpy) is a Linear Algebra Library for Python, the reason it is so important for Data Science with Python is that almost all of the libraries in the PyData Ecosystem rely on NumPy as one of their main building blocks.
  
−
* NumPy is also incredibly fast, as it has bindings to C libraries.
+
* Numpy is also incredibly fast, as it has bindings to C libraries. For more info on why you would want to use Arrays instead of lists, check out this great [StackOverflow post](http://stackoverflow.com/questions/993984/why-numpy-instead-of-python-lists).
−
 
 
−
* It is highly recommended you install Python using the Anaconda distribution to make sure all anderlying dependencies (such as Linear Algebra libraries) all sync up with the use of a conda install.
 
  
  
 
<br />
 
<br />
 
====Installation====
 
====Installation====
 +
It is highly recommended you install Python using the Anaconda distribution to make sure all underlying dependencies (such as Linear Algebra libraries) all sync up with the use of a conda install.
 +
 
* If you have Anaconda, install NumPy by:
 
* If you have Anaconda, install NumPy by:
 
  conda install numpy
 
  conda install numpy
Line 106: Line 106:
  
  
−
Let's create a List. Then, using numpy we can cast this List into an Array. In this case, a 1-d Array (Vector)
+
<br />
−
<syntaxhighlight lang="python">
+
=====Creating NumPy Arrays=====
 +
 
 +
 
 +
<br />
 +
======From a Python List======
 +
We can create an array by directly converting a list or list of lists:
 
my_list = [1,2,3]
 
my_list = [1,2,3]
 +
np.array(my_list)
  
−
import numpy as np
+
my_matrix = [[1,2,3],[4,5,6],[7,8,9]]
 +
np.array(my_matrix)
  
−
vector = np.array(my_list)
 
−
</syntaxhighlight>
 
  
 +
<br />
 +
======From NumPy Built-in Methods======
 +
Return evenly spaced values within a given interval.
  
−
If we want a 2-d Array, we can first create a List of List and then we can cast it into a 2-d Matrix:
+
'''arange'''
−
<syntaxhighlight lang="python">
+
np.arange(0,10)
−
list_of_lists = [[1,2,3],[4,5,6],[7,8,9]]
+
np.arange(0,11,2)
−
</syntaxhighlight>
 
  
  
−
Creating an Array using NumPy built-in generation methods:
+
'''zeros and ones''''
−
<syntaxhighlight lang="python">
 
−
np.arange(0,10)
 
  
−
np.zeros(10)
+
zeros and ones
−
np.zeros((2,3)) # In this example: N. of rows: 2 ; N. of columns: 3
 
  
−
np.ones((3,4))
+
np.zeros((5,5))
−
</syntaxhighlight>
 
  
 +
np.ones(3)
 +
np.ones((3,3))
  
−
Lucture 18, min 4:20
+
 
−
<br />
+
'''linspace'''
 +
 
 +
Return evenly spaced numbers over a specified interval.
 +
 
 +
np.linspace(0,10,3)
 +
 
 +
np.linspace(0,10,50)
 +
 
 +
 
 +
'''eye'''
 +
Creates an identity matrix
 +
 
 +
np.eye(4)
 +
 
 +
 
 +
'''Random'''
 +
Numpy also has lots of ways to create random number arrays:
 +
 
 +
rand
 +
 
 +
Create an array of the given shape and populate it with random samples from a uniform distribution over [0, 1).

Revision as of 23:19, 15 October 2019

Anaconda

Anaconda is a free and open source distribution of the Python and R programming languages for data science and machine learning related applications (large-scale data processing, predictive analytics, scientific computing), that aims to simplify package management and deployment. Package versions are managed by the package management system conda. https://en.wikipedia.org/wiki/Anaconda_(Python_distribution)

En otras palabras, Anaconda puede ser visto como un paquete (a distribution) que incluye no solo Python (or R) but many libraries that are used in Data Science, as well as its own virtual environment system. It's an "all-in-one" install that is extremely popular in data science and Machine Learning.



Installation

https://linuxize.com/post/how-to-install-anaconda-on-ubuntu-18-04/

https://www.digitalocean.com/community/tutorials/how-to-install-the-anaconda-python-distribution-on-ubuntu-18-04



Anaconda comes with a few IDE

  • Jupyter Lab
  • Jupyter Notebook
  • Spyder
  • Qtconsole
  • and others



Anaconda Navigator

Anaconda Navigator is a GUI that helps you to easily start important applications and manage the packages in your local Anaconda installation

You can open the Anaconda Navigator from the Terminal: 

anaconda-navigator



Jupyter

Jupyter comes with Anaconda.

  • It is a development environment (IDE) where we can write codes; but it also allows us to display images, and write down markdown notes.
  • It is the most popular IDE in data science for exploring and analyzing data.
  • Other famoues IDE for Python are Sublime Text and PyCharm.
  • There is Jupyter Lab and Jupyter Notebook



Online Jupyter

There are many sites that provides solutions to run your Jupyter Notebook in the cloud: https://www.dataschool.io/cloud-services-for-jupyter-notebook/

I have tried:

https://cocalc.com/projects/595bf475-61a7-47fa-af69-ba804c3f23f9/files/?session=default
Parece bueno, pero tiene opciones que no son gratis


https://www.kaggle.com/adeloaleman/kernel1917a91630/edit
Parece bueno pero no encontré la forma adicionar una TOC


Es el que estoy utilizando ahora



Udemy - Python for Data Science and Machine Learning Bootcamp

https://www.udemy.com/course/python-for-data-science-and-machine-learning-bootcamp/



Most popular Python Data Science Libraries

  • NumPy
  • SciPy
  • Pandas
  • Seaborn
  • SciKit'Learn
  • MatplotLib
  • Plotly
  • PySpartk



NumPy

  • NumPy (or Numpy) is a Linear Algebra Library for Python, the reason it is so important for Data Science with Python is that almost all of the libraries in the PyData Ecosystem rely on NumPy as one of their main building blocks.



Installation

It is highly recommended you install Python using the Anaconda distribution to make sure all underlying dependencies (such as Linear Algebra libraries) all sync up with the use of a conda install.

  • If you have Anaconda, install NumPy by:
conda install numpy
  • If you are not using Anaconda distribution:
pip install numpy



NumPy Arrays

  • NumPy arrays are the main way we will use NumPy throughout the course.
  • NumPy arrays essentially come in two flavors: vectors and matrices.
  • Vectors are strictly 1-d arrays and matrices are 2-d (but you should note a matrix can still have only one row or one column).



Creating NumPy Arrays


From a Python List

We can create an array by directly converting a list or list of lists: my_list = [1,2,3] np.array(my_list)

my_matrix = [[1,2,3],[4,5,6],[7,8,9]] np.array(my_matrix)



From NumPy Built-in Methods

Return evenly spaced values within a given interval.

arange np.arange(0,10) np.arange(0,11,2)


zeros and ones'

zeros and ones

np.zeros((5,5))

np.ones(3) np.ones((3,3))


linspace

Return evenly spaced numbers over a specified interval.

np.linspace(0,10,3)

np.linspace(0,10,50)


eye Creates an identity matrix

np.eye(4)


Random Numpy also has lots of ways to create random number arrays:

rand

Create an array of the given shape and populate it with random samples from a uniform distribution over [0, 1).

Retrieved from "http://wiki.sinfronteras.ws/index.php?title=Python_for_Data_Science&oldid=11972"