Difference between revisions of "Python for Data Science"

From Sinfronteras
Jump to: navigation, search
(Keep a python script running on a remote server)
(Replaced content with "~ Migrated")
(Tag: Replaced)
 
(9 intermediate revisions by the same user not shown)
Line 1: Line 1:
 
+
~ Migrated
<br />
 
For a standard Python tutorial go to [[Python]]
 
 
 
 
 
<br />
 
==Courses==
 
*Udemy - Python for Data Science and Machine Learning Bootcamp
 
 
 
:https://www.udemy.com/course/python-for-data-science-and-machine-learning-bootcamp/
 
 
 
 
 
<br />
 
==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.Creating sample array for the following examples:
 
 
 
 
 
<br />
 
===Installation===
 
Installation from the official Anaconda Web site: https://docs.anaconda.com/anaconda/install/
 
 
 
 
 
<br />
 
 
 
===Anaconda comes with a few IDE===
 
 
 
*Jupyter Lab
 
*Jupyter Notebook
 
*Spyder
 
*Qtconsole
 
*and others
 
 
 
 
 
<br />
 
===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
 
 
 
 
 
<br />
 
 
 
==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'''
 
 
 
 
 
<br />
 
===Remote connection===
 
https://jupyter-notebook.readthedocs.io/en/stable/public_server.html
 
 
 
 
 
A**1
 
 
 
 
 
<syntaxhighlight lang="shell">
 
(base) adelo@vmi346715:~/.jupyter$ openssl req -x509 -nodes -days 365 -newkey rsa:2048 -keyout mykey.key -out mycert.pem
 
Generating a RSA private key
 
......................................+++++
 
....................................+++++
 
writing new private key to 'mykey.key'
 
-----
 
You are about to be asked to enter information that will be incorporated
 
into your certificate request.
 
What you are about to enter is what is called a Distinguished Name or a DN.
 
There are quite a few fields but you can leave some blank
 
For some fields there will be a default value,
 
If you enter '.', the field will be left blank.
 
-----
 
Country Name (2 letter code) [AU]:IE
 
State or Province Name (full name) [Some-State]:Dublin
 
Locality Name (eg, city) []:Dublin
 
Organization Name (eg, company) [Internet Widgits Pty Ltd]:.
 
Organizational Unit Name (eg, section) []:.
 
Common Name (e.g. server FQDN or YOUR name) []:sinfronteras   
 
Email Address []:adeloaleman@gmail.com
 
</syntaxhighlight>
 
 
 
 
 
<br />
 
===Share Jupyter Notebook online===
 
* '''GitHub:'''
 
: https://docs.github.com/en/github/managing-files-in-a-repository/working-with-jupyter-notebook-files-on-github
 
: Example: https://github.com/adeloaleman/AmazonLaptopsDashboard/blob/master/DataAnalysis/data_analysis2.ipynb
 
 
 
 
 
* '''Nbviewer''
 
: https://nbviewer.jupyter.org/
 
: Example: https://nbviewer.jupyter.org/github/bokeh/bokeh-notebooks/blob/main/tutorial/06%20-%20Linking%20and%20Interactions.ipynb
 
 
 
 
 
<br />
 
 
 
===Customize Jupyter===
 
 
 
 
 
<br />
 
====Themes====
 
https://github.com/dunovank/jupyter-themes
 
 
 
Ver el tema que muestran en esta página: https://gist.github.com/pierrejoubert73/902cc94d79424356a8d20be2b382e1ab
 
 
 
 
 
jt  -t oceans16    -cellw 98%  -lineh 120  -fs 14  -nfs 14  -dfs 14  -ofs 14
 
 
 
 
 
https://www.kaggle.com/getting-started/97540
 
jt  -t monokai      -cellw 98%  -lineh 120  -fs 14  -nfs 14  -dfs 14  -ofs 14  -f fira  -nf ptsans  -N  -kl  -cursw 2  -cursc r  -T
 
 
 
 
 
<br />
 
 
 
====Extensions====
 
This post mention so nice extension and configuration that can be done: https://towardsdatascience.com/bringing-the-best-out-of-jupyter-notebooks-for-data-science-f0871519ca29
 
 
 
<br />
 
=====Unofficial Jupyter Notebook Extensions=====
 
https://jupyter-contrib-nbextensions.readthedocs.io/en/latest/index.html
 
 
 
<span style="color: green">'''This is very important. There are very nice extensions in this package:'''</span>
 
 
 
* toc2: https://jupyter-contrib-nbextensions.readthedocs.io/en/latest/nbextensions/toc2/README.html
 
* Collapsible Headings
 
* ... etc
 
 
 
<br />
 
======Installation======
 
https://jupyter-contrib-nbextensions.readthedocs.io/en/latest/install.html
 
 
 
<span style="color: red">'''I had some issues to install it. La format indicada por defecto:'''</span>
 
 
 
pip install jupyter_contrib_nbextensions
 
jupyter contrib nbextension install --user
 
 
 
<span style="color: red">'''A través de la forma anterior no pude instalar el paquete de forma correcta. La instalación no retornó errorres, y la extensión se mostraba en Jupyter-notebook pero no podía activar "enable" las extensiones.'''</span>
 
 
 
 
 
<span style="color: red">'''Al parecer es un problema con la ubicación de la instalación. Yo estaba usando conda pero conda está presentando problemas. La instalación de los paquestes demora muchísimo y luego el paquete parece no estar disponible.'''</span>
 
 
 
 
 
<span style="color: red">'''En el siguiente post encontré una solución para instalar nbextension usando pip:'''</span>
 
https://github.com/ipython-contrib/jupyter_contrib_nbextensions/issues/1127
 
 
 
pip install --upgrade jupyter_contrib_nbextensions
 
jupyter contrib nbextension install  --sys-prefix  --symlink
 
 
 
<span style="color: red">'''«--symlink» creo que lo usé pero no estoy completamente seguro. También realicé el --upgrade pero creo que la diferencia la hicieron las opciones --sys-prefix  --symlink'''</span>
 
 
 
 
 
 
 
Si no se muestra la '''Nbextensions''' tab (), try to reinstall the https://github.com/Jupyter-contrib/jupyter_nbextensions_configurator
 
 
 
pip install jupyter_nbextensions_configurator
 
or
 
conda install -c conda-forge jupyter_nbextensions_configurator
 
 
 
 
 
<br />
 
 
 
====CustomJS and CustonCSS files====
 
This is a good post: https://forums.fast.ai/t/jupyter-notebook-enhancements-tips-and-tricks/17064
 
 
 
Keyboard Shortcut Customization: https://jupyter-notebook.readthedocs.io/en/stable/examples/Notebook/Custom%20Keyboard%20Shortcuts.html
 
 
 
 
 
<br />
 
custom.js
 
<syntaxhighlight lang="js">
 
/** Mis configuraciones */
 
 
 
// This is to enable syntax highlighting for SQL code:
 
// https://stackoverflow.com/questions/43641362/adding-syntax-highlighting-to-jupyter-notebook-cell-magic
 
require(['notebook/js/codecell'], function(codecell) {
 
  codecell.CodeCell.options_default.highlight_modes['magic_text/x-mssql'] = {'reg':[/^%%sql/]} ;
 
  Jupyter.notebook.events.one('kernel_ready.Kernel', function(){
 
  Jupyter.notebook.get_cells().map(function(cell){
 
      if (cell.cell_type == 'code'){ cell.auto_highlight(); } }) ;
 
  });
 
});
 
 
 
 
 
// My plain theme
 
// This is a good post where I took some ideas to write the following fuction: https://forums.fast.ai/t/jupyter-notebook-enhancements-tips-and-tricks/17064
 
function plainTheme() {
 
    var input_promp_fields = document.getElementsByClassName("prompt_container");
 
    var text_render_fields = document.getElementsByClassName("text_cell_render");
 
 
 
    if (input_promp_fields[0].style.visibility == "collapse"){
 
        action = "visible";
 
        input_marginLeft = "0px";
 
        border_top  = "3px";
 
        prompt_width = "74px";
 
        padding_top = "0px";
 
        output_margin = "40px";
 
    }else{
 
        action = "collapse";
 
        input_marginLeft = "74px";
 
        border_top  = '0px';
 
        prompt_width = "74px";
 
        padding_top = "40px";
 
        output_margin = "40px";
 
    }
 
 
 
    // Si queremos usar !important debemos hacerlo de esta forma utilizando JQuery:
 
    // https://makitweb.com/how-to-add-important-to-css-property-with-jquery/
 
    var text_cell_fields = document.getElementsByClassName("text_cell");
 
    $(text_cell_fields).ready(function(){
 
        $('.input_prompt').css({
 
            'cssText': `width: 40px !important; max-width: ${prompt_width} !important; min-width: ${prompt_width} !important;`
 
        });
 
    });
 
 
 
    $(document).ready(function(){
 
        $(".prompt_container").css(
 
            'visibility', `${action}`
 
        );
 
       
 
        $(".input").css(
 
            'padding-left', `${input_marginLeft}`
 
        );
 
       
 
        $(".output_subarea").css(
 
            'margin-left', `${output_margin}`
 
        );
 
                   
 
        $('.cell').css({
 
            'cssText': `border-top-width: ${border_top} !important; border-bottom-width: ${border_top} !important;`
 
        });
 
       
 
        $(".collapsible_headings_ellipsis").css({
 
            'cssText': `padding-top:${padding_top} !important; border-top-width: ${border_top} !important; border-bottom-width: ${border_top} !important;`
 
        });
 
 
 
        $(".text_cell_render").css({
 
            'cssText': `margin-left: -10px;`
 
        });
 
    });           
 
}
 
 
 
Jupyter.keyboard_manager.command_shortcuts.add_shortcut('Alt-Ctrl-Q', {
 
    help : '...',
 
    help_index : 'zz',
 
    handler : function (event) {
 
        plainTheme();
 
    return false;
 
    }}
 
);
 
 
 
Jupyter.keyboard_manager.edit_shortcuts.add_shortcut('Alt-Ctrl-Q', {
 
    help : '...',
 
    help_index : 'zz',
 
    handler : function (event) {
 
        plainTheme();
 
    return false;
 
    }}
 
);
 
 
 
 
 
// This could be very usefull. It allows to add text automatically into a cell
 
// https://forums.fast.ai/t/jupyter-notebook-enhancements-tips-and-tricks/17064/27
 
Jupyter.keyboard_manager.edit_shortcuts.add_shortcut('Ctrl-Shift-J', {
 
    help : '...',
 
    help_index : 'zz',
 
    handler : function (event) {
 
        document.body.style.background = 'blue'
 
        var target = Jupyter.notebook.get_selected_cell()
 
        var cursor = target.code_mirror.getCursor()
 
        var before = target.get_pre_cursor()
 
        var after = target.get_post_cursor()
 
        target.set_text(before + 'from IPython.core.display import display, HTML; \n\taverrrdisplay(HTML("<style>.container { width:98% !important;}</style>"))' + after)
 
        cursor.ch += 20 // where to put your cursor
 
        target.code_mirror.setCursor(cursor)
 
        return false;
 
    }}
 
);
 
 
 
 
 
// To get the real value of a css field: https://stackoverflow.com/questions/26074476/document-body-style-backgroundcolor-doesnt-work-with-external-css-style-sheet
 
// window.getComputedStyle(document.body).backgroundColor
 
// window.getComputedStyle(document.getElementsByClassName("input_area")[0]).backgroundColor
 
</syntaxhighlight>
 
 
 
 
 
<br />
 
custom.css
 
<syntaxhighlight lang="css">
 
/*  Mis configuraciones  */
 
 
 
.container { width:98% !important; }
 
/* document.getElementById("notebook-container").style.minWidth = "50%"; */
 
/* document.getElementById("notebook-container").style.maxWidth = "50%"; */
 
 
 
#notebook-container {
 
width:98% !important;
 
}
 
 
 
.CodeMirror-gutters {
 
background-color: transparent !important;
 
background: transparent !important;
 
}
 
 
 
.CodeMirror-linenumber {
 
margin-left: -20px !important;
 
}
 
 
 
.output_subarea {
 
margin-left: 40px !important;
 
}
 
 
 
#toc .fa-fw {
 
color: blue !important;
 
}
 
 
 
#toc .highlight_on_scroll {
 
margin-left: -4px !important;
 
 
}
 
 
 
#toc {
 
padding-left: 10px !important;
 
}
 
 
 
/*  I have also changed the color
 
/*  #a6e22e  by  #388bfd
 
*  in the entire custom.css
 
*/
 
 
 
/* I have also chenged some of the properties of the toc directly above in the code:
 
 
 
#toc-wrapper {
 
z-index: 90;
 
position: fixed !important;
 
display: flex;
 
flex-direction: column;
 
overflow: hidden;
 
padding: 10px;
 
padding-top: 40px !important;
 
border-style: solid;
 
border-width: thin;
 
border-right-width: medium !important;
 
background-color: #1e1e1e !important;
 
}
 
#toc-wrapper.ui-draggable.ui-resizable.sidebar-wrapper {
 
border-color: rgba(93,92,82,.25) !important;
 
}
 
#toc a,
 
#navigate_menu a,
 
.toc {
 
color: #f8f8f0 !important;
 
font-size: 16pt !important;
 
}
 
#toc li > span:hover {
 
background-color: rgba(93,92,82,.25) !important;
 
}
 
#toc a:hover,
 
#navigate_menu a:hover,
 
.toc {
 
color: #DAA520 !important;
 
font-size: 16pt !important;
 
}
 
#toc-wrapper .toc-item-num {
 
color: #388bfd !important;
 
font-size: 16pt !important;
 
}
 
*/
 
</syntaxhighlight>
 
 
 
 
 
<br />
 
 
 
====Configurations from the Juniper notebook====
 
 
 
<syntaxhighlight lang="python3">
 
from IPython.core.display import display, HTML;
 
 
 
display(HTML("<style>.container { width:98% !important;}</style>"<))
 
 
 
display(HTML('<style>.prompt.input_prompt{display:none !important;}</style>'))
 
display(HTML('<style>.prompt.input_prompt{visibility: visible !important;</style>'))
 
display(HTML('<style>.prompt.input_prompt{margin-left8kmclustering.ipynb 50px}</style>'))
 
display(HTML('<style>.prompt.input_prompt{visibility: visible !important; width: 0px !important; min-width: 0px !important}</style>')) 
 
 
 
display(HTML('<style>.input_area{margin-left: -50px;}</style>'))
 
display(HTML('<style>.input{margin-left: -20px;}</style>'))
 
 
 
display(HTML('<style>.output_area{margin-left: 55px}</style>'))
 
 
 
# display(HTML('<style>.cell{margin-bottom: -5px !important; margin-top: -5px !important;}</style>'))
 
# display(HTML('<style>.code_cell{margin-bottom: -5px !important; margin-top: -5px !important;}</style>'))
 
 
 
# display(HTML('<style>.output_wrapper{margin-bottom: 0px !important; margin-top: 0px !important;}</style>'))
 
</syntaxhighlight>
 
 
 
 
 
<br />
 
 
 
===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/app
 
 
 
::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/
 
 
 
::https://www.kaggle.com/adeloaleman/kernel1917a91630/edit
 
::Parece bueno pero no encontré la forma adicionar una TOC
 
 
 
 
 
*https://drive.google.com
 
 
 
:*https://colab.research.google.com
 
::Es el que estoy utilizando ahora
 
 
 
 
 
<br />
 
===Some remarks===
 
 
 
 
 
<br />
 
====Executing Terminal Commands in Jupyter Notebooks====
 
https://support.anaconda.com/hc/en-us/articles/360023858254-Executing-Terminal-Commands-in-Jupyter-Notebooks
 
 
 
If we are in the Notebook, and we want to run a shell command rather than a notebook command we use the <code>'''!''' or '''%'''</code>
 
 
 
Try, for example:
 
%ls
 
!pwd
 
 
 
It's the same as if you opened up a terminal and typed it without the <code>'''!'''</code>
 
 
 
 
 
<br />
 
 
 
===[[HTML presentation with Reveal.js#Creating Presentations in Jupyter Notebook with RevealJS|Creating Presentations in Jupyter Notebook with RevealJS]]===
 
 
 
 
 
<br />
 
 
 
==Some of the most popular Python Data Science Libraries==
 
 
 
*NumPy
 
*SciPy
 
*Pandas
 
*Seaborn
 
*SciKit'Learn
 
*MatplotLib
 
*Plotly
 
*PySpartk
 
 
 
 
 
<br />
 
 
 
==[[NumPy and Pandas]]==
 
 
 
 
 
<br />
 
==[[Data Visualization with Python]]==
 
 
 
 
 
<br />
 
 
 
==[[Natural Language Processing]]==
 
 
 
 
 
<br />
 
==[[Dash - Plotly]]==
 
 
 
 
 
<br />
 
==[[Scrapy]]==
 
 
 
 
 
<br />
 
==Using SQL in Jupyter==
 
Connecting to a database in Jupyter
 
 
 
 
 
https://pypi.org/project/ipython-sql/
 
 
 
https://stackoverflow.com/questions/454854/no-module-named-mysqldb
 
 
 
https://stackoverflow.com/questions/5178292/pip-install-mysql-python-fails-with-environmenterror-mysql-config-not-found
 
 
 
https://docs.kyso.io/guides/sql-interface-within-jupyterlab
 
 
 
https://www.datacamp.com/community/tutorials/sql-interface-within-jupyterlab
 
 
 
https://stackoverflow.com/questions/43641362/adding-syntax-highlighting-to-jupyter-notebook-cell-magic
 
 
 
https://www.sqlshack.com/learn-jupyter-notebooks-for-sql-server/
 
 
 
 
 
Verificar las fuentes above. Creo que lo único que tuve que hacer la última vez que lo instalé fue basado en las 3 primeras sources:
 
 
 
pip install ipython-sql
 
 
sudo apt install default-libmysqlclient-dev
 
 
pip install mysqlclient
 
 
sudo apt-get install python3-mysqldb
 
 
 
 
 
Luego adding SQL syntax highlighting to Jupyter as describe above in the corrrespoinding source.
 
 
 
 
 
<br />
 
 
 
==Keep a python script running on a remote server==
 
 
 
 
 
<br />
 
===Screen===
 
 
 
Create a screen:
 
screen -S bot
 
 
 
Disconnect: Ctrl+A+D
 
 
 
Reconnecting:
 
screen -r bot
 
 
 
 
 
List screens:
 
screen -ls
 
 
 
 
 
Kill all sessions
 
killall screen
 
 
 
 
 
Creating the screen without attaching to it
 
screen -dmS bot
 
 
 
 
 
Activating the venv on the screen previously created without attaching to it
 
screen -S bot -p 0 -X stuff $'source .venv/bin/activate\n'
 
 
 
 
 
To create the screen and activate the venv at the same time without attaching to it:
 
screen -dmS bot bash -c 'source .venv/bin/activate && exec sh'    # This is working but the prompt is sh so not very good. (&& can be replaced by ;)
 
screen -dmS bot bash -c 'source .venv/bin/activate && exec bash'  # This is not activating the venv
 
 
 
 
 
<br />
 

Latest revision as of 11:33, 28 February 2026

~ Migrated