Notebook Best Practices

Use Collapsible Headings and Table of Content
Notebooks should be executable from top to bottom
Name your variables carefully
Use dummy names such as tmp or _ when needed
Clear useless variables when not needed (del my_variable)
Clear your code and merge cells when relevant (Shift-M)
Hide your cell outputs to gain space (double-click on the red Out[]: section to the left of your cell).

from IPython.display import HTML, IFrame
IFrame("http://www.youtube.com/embed/8QiPFmIMxFc?t=388", width="560", height="315")

from ipywidgets import interact

@interact
def plot_polynom(a=[0,1,2,3], b=2):
    x = np.arange(-10, 10, 0.1)
    y = a*x**3+ b*x**2    
    plt.plot(x,y); plt.xlim(xmin=-10, xmax=10); plt.ylim(ymin=-100, ymax=100)

from ipywidgets import interact

@interact
def plot_polynom(a=[0,1,2,3], b=2):
    x = np.arange(-10, 10, 0.1)
    y = a*x**3+ b*x**2    
    plt.plot(x,y); plt.xlim(xmin=-10, xmax=10); plt.ylim(ymin=-100, ymax=100)

# Importing Libraries
import matplotlib.pyplot as plt
import pandas as pd
import numpy as np
import seaborn as sns

# Dataprep for exploratory data analysis
from dataprep.eda import create_report
from dataprep.eda import plot, plot_correlation, plot_missing

# mport Regressor Metric Graph Plot - for ML analysis
from regressormetricgraphplot import *

# Set style to 'seaborn / Plot inline
plt.style.use('seaborn')
%matplotlib inline
%config InlineBackend.figure_format = 'retina'


# Load Dataframe
df = pd.read_csv('path/to/file.csv')
df.head()

df.isna().sum()
df.drop_duplicates(inplace = True)
df.info()
df.describe()

# Generate Dataprep Report
report = create_report(df, title='My Report')
create_report()

plot(df, col1, col2)

# df sortby
df.sort_values(by=[col1, col2], ascending=[True, True])

# df groupby
df.groupby(col).mean()

cat_type = CategoricalDtype(categories=['3', '2', '1'], ordered=True)
cat_type2 = CategoricalDtype(categories=['Kind','Jong','Middelbaar','Oud'], ordered=True)
​
df1['pclass'] = df1['pclass'].map({1: '1', 2: '2', 3: '3'})
df1['survived'] = df1['survived'].map({1: True, 0: False})
df1['sex'] = df1['sex'].map({'male': 'M', 'female': 'V'})
df1['pclass'] = df1['pclass'].astype(cat_type)
df1['sex'] = df1['sex'].astype('category')
df1['age'] = df1['age'].map(lambda x: round(x))
df1['age'] = df1['age'].astype('int8')
df1['fare'] = df1['fare'].map(lambda x: round(x, 2))
df1['age_cat'] = pd.cut(df1['age'], bins=4, labels=('Kind','Jong','Middelbaar','Oud'))
df1['age_cat'] = df1['age_cat'].astype(cat_type2)
df1 = df1.filter(items=['pclass', 'name', 'survived', 'sex', 'age', 'age_cat', 'fare'])
df1

<div style="background-color:rgb(250,250,250); padding:30px;box-shadow:0 1px 1px rgba(0, 0, 0, 0.2);">
    <p style="font-family:Helvetica;font-size:15px;font-weight:bold">
       Text....
    </p>
</div>

<div style="background-color:#1abc9c; padding:1px 20px 20px 20px; border-radius:5px;box-shadow: rgba(0, 0, 0, 0.25) 0px 0.0625em 0.0625em, rgba(0, 0, 0, 0.25) 0px 0.125em 0.5em, rgba(255, 255, 255, 0.1) 0px 0px 0px 1px inset;">
    <h1 style="font-family: 'Lato', sans-serif;color:white;padding-left:10px;font-size:50px">
        Heading
    </h1>
</div>

<div class="alert alert-block alert-info" style="border-radius:8px; box-shadow: rgba(0, 0, 0, 0.1) 0px 1px 3px 0px, rgba(0, 0, 0, 0.06) 0px 1px 2px 0px;">
    <b>&#9432; Note<br /></b> Use blue boxes (alert-info) for tips and notes.
    If it’s a note, you don’t have to include the word “Note”.
</div>

<div class="alert alert-block alert-warning" style="border-radius:8px; box-shadow: rgba(0, 0, 0, 0.1) 0px 1px 3px 0px, rgba(0, 0, 0, 0.06) 0px 1px 2px 0px;">
    <b>!&#x20DD; Important<br /></b> Use yellow boxes if you to underline important things.
</div>

<div class="alert alert-block alert-danger" style="border-radius:8px; box-shadow: rgba(0, 0, 0, 0.1) 0px 1px 3px 0px, rgba(0, 0, 0, 0.06) 0px 1px 2px 0px;">
    <b>&#9888; Warning<br /></b> In general, avoid the red boxes. These should only be
    used for actions that might cause data loss or another major issue.
</div>

<div class="alert alert-block alert-success" style="border-radius:8px; box-shadow: rgba(0, 0, 0, 0.1) 0px 1px 3px 0px, rgba(0, 0, 0, 0.06) 0px 1px 2px 0px;">
    <b>&#10149; See also<br /></b> Use green boxes to link to other documentation sources.
</div>

conda create -n my-conda-env                               # creates new virtual env
conda activate my-conda-env                                # activate environment in terminal
conda install ipykernel                                    # install Python kernel in new conda env
ipython kernel install --user --name=my-conda-env-kernel   # configure Jupyter to use Python kernel
jupyter notebook                                           # run jupyter from system

(firstEnv)
>>conda install -c anaconda ipykernel
>>python -m ipykernel install --user --name=firstEnv

#!/usr/bin/env python3

from multiprocessing import Pool

def run(task):
  # Do something with task here
    print("Handling {}".format(task))

if __name__ == "__main__":
  tasks = ['task1', 'task2', 'task3']
  # Create a pool of specific number of CPUs
  p = Pool(len(tasks))
  # Start each task within the pool
  p.map(run, tasks)

# this line will write the code below into a Python script called script.py
%%writefile script.py