Section 1.1: Core Libraries

1. Pandas

   The first library on our list is Pandas, an indispensable tool for data scientists. It offers high-level data structures and manipulation capabilities that simplify data analysis.

import pandas as pd

# Creating a DataFrame

df = pd.DataFrame({

'A': ['foo', 'bar', 'foo', 'bar', 'foo', 'bar', 'foo', 'foo'],

'B': ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'],

'C': np.random.randn(8),

'D': np.random.randn(8)

})

print(df)

2. Numpy

   Next is Numpy, a library that provides support for large multi-dimensional arrays and matrices, along with a suite of mathematical functions to operate on them.

import numpy as np

# Create an array and perform operations

a = np.array([1, 2, 3])

b = np.array([4, 5, 6])

print("Array sum: ", a + b)

3. Matplotlib

   Visualization is crucial in data science, and Matplotlib serves as a robust tool for creating static, animated, and interactive plots.

import matplotlib.pyplot as plt

# Sample plot

plt.plot([1, 2, 3, 4])

plt.ylabel('some numbers')

plt.show()

4. Scikit-learn

   Scikit-learn is a machine learning library that offers various classification, regression, and clustering algorithms, built on top of Numpy and Matplotlib.

from sklearn import svm, datasets

# Load dataset and create a model

iris = datasets.load_iris()

X, y = iris.data, iris.target

clf = svm.SVC()

clf.fit(X, y)

5. TensorFlow

   Developed by Google, TensorFlow is a library for efficient numerical computing and serves as a foundation for building and training machine learning models.

import tensorflow as tf

# A simple computation in TensorFlow

a = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[2, 3])

b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[3, 2])

c = tf.matmul(a, b)

print(c)

6. Seaborn

   Seaborn builds on Matplotlib and offers a high-level interface for drawing attractive and informative statistical graphics.

import seaborn as sns

# Load the iris dataset

iris = sns.load_dataset("iris")

# Construct iris plot

sns.swarmplot(x="species", y="petal_length", data=iris)

7. Keras

   Keras is an open-source neural network library that is user-friendly and modular, making it easier to create neural networks.

from keras.models import Sequential

from keras.layers import Dense

# Define a simple model

model = Sequential()

model.add(Dense(12, input_dim=8, activation='relu'))

model.add(Dense(8, activation='relu'))

model.add(Dense(1, activation='sigmoid'))

8. NLTK

   The Natural Language Toolkit (NLTK) is essential for those working with natural language processing (NLP), providing interfaces to over 50 corpora and lexical resources.

import nltk

# Tokenize a sentence

from nltk.tokenize import word_tokenize

print(word_tokenize("Hello, world!"))

9. SciPy

   SciPy is an open-source library designed for scientific and technical computing, building on Numpy and offering numerous higher-level scientific algorithms.

from scipy import linalg, sparse

# Create a 2D array

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

# Perform operations using linalg

print(linalg.det(A))

10. PyTorch

    PyTorch is an open-source machine learning library developed by Facebook's AI Research lab, widely used for applications like natural language processing.

import torch

# Create tensors

x = torch.tensor([1.0])

y = torch.tensor([2.0])

# Multiply tensors

z = x * y

print(z)

Section 1.2: Lesser-Known Libraries

Now, let's explore some intriguing yet lesser-known Python libraries for data science.

11. Dask

    Dask is a flexible library for parallel computing, designed with the core Python data science stack in mind.

import dask.array as da

# Create a large random array in chunks

x = da.random.random((10000, 10000), chunks=(1000, 1000))

# Compute and return the mean

print(x.mean().compute())

12. Yellowbrick

    Yellowbrick enhances the Scikit-learn API, making model selection and hyperparameter tuning more accessible.

from yellowbrick.datasets import load_energy

from yellowbrick.target import BalancedBinningReference

# Load a regression dataset

X, y = load_energy()

# Instantiate the visualizer

visualizer = BalancedBinningReference()

visualizer.fit(y)

visualizer.show()

13. Eli5

    Eli5 is a library that helps debug machine learning classifiers and interpret their predictions, supporting many popular libraries.

from sklearn import datasets

from sklearn.ensemble import RandomForestClassifier

import eli5

# Training a classifier

iris = datasets.load_iris()

X, y = iris.data, iris.target

clf = RandomForestClassifier(random_state=42)

clf.fit(X, y)

# Explaining weights

print(eli5.explain_weights(clf))

14. PyCaret

    PyCaret is a low-code machine learning library that automates workflows in Python, offering an end-to-end ML solution.

from pycaret.datasets import get_data

from pycaret.classification import *

# Get a dataset

diabetes = get_data('diabetes')

# Setup ML Experiment

exp = setup(data=diabetes, target='Class variable')

# Compare models

compare_models()

15. Imbalanced-learn

    Imbalanced-learn is a Python library designed to address imbalanced datasets, compatible with Scikit-learn.

from imblearn.over_sampling import RandomOverSampler

ros = RandomOverSampler(random_state=0)

X_resampled, y_resampled = ros.fit_resample(X, y)

print(sorted(Counter(y_resampled).items()))