Skip to main content

Libraries

Scientific Computing Libraries in Python​

Pandas​

  • Purpose: Data structures and tools for effective data cleaning, manipulation, and analysis.
  • Key Features:
    • Primary instrument: Data Frame (two-dimensional table of columns and rows).
    • Easy indexing for data manipulation.

NumPy​

  • Purpose: Array and matrix operations.
  • Key Features:
    • Mathematical functions on arrays.
    • Foundation for many other libraries, including Pandas.

Visualization Libraries in Python​

Matplotlib​

  • Purpose: Creating graphs and plots.
  • Key Features:
    • Customizable graphs.
    • Widely used for a variety of visualizations.

Seaborn​

  • Purpose: High-level interface for drawing attractive statistical graphics.
  • Key Features:
    • Based on Matplotlib.
    • Generates heat maps, time series, violin plots, etc.

High-Level Machine Learning and Deep Learning Libraries in Python​

Scikit-learn​

  • Purpose: Tools for statistical modeling, including regression, classification, and clustering.
  • Key Features:
    • Built on NumPy, SciPy, and Matplotlib.
    • Simple to get started with defining models and specifying parameters.

Keras​

  • Purpose: Building standard deep learning models quickly and easily.
  • Key Features:
    • High-level interface.
    • Can use GPUs for processing.

Deep Learning Libraries in Python​

TensorFlow​

  • Purpose: Production and deployment of large-scale deep learning models.
  • Key Features:
    • Low-level framework.
    • Suitable for large-scale production.

PyTorch​

  • Purpose: Experimentation in deep learning research.
  • Key Features:
    • Simple for researchers to test ideas.

Libraries Used in Other Languages​

Apache Spark​

  • Purpose: General-purpose cluster-computing framework.
  • Key Features:
    • Processes data using compute clusters.
    • Similar functionality to Pandas, NumPy, and Scikit-learn.
    • Data processing jobs can be in Python, R, Scala, and SQL.

Scala Libraries​

  • Vegas: Statistical data visualizations.
    • Works with data files and Spark Data Frames.
  • Big DL: Deep learning library.

R Libraries​

  • ggplot2: Data visualization.
  • Libraries for interfacing with Keras and TensorFlow.
  • Built-in functionality for machine learning and data visualization.

Summary​

  • Libraries provide built-in modules for various functionalities.
  • Data visualization methods are essential for communicating analysis results.
  • Scikit-learn offers tools for statistical modeling in machine learning.
  • TensorFlow is used for large-scale production of deep learning models.
  • Apache Spark processes data using compute clusters and supports multiple languages.