Doing more with Python Numpy

Tap full potential of Numpy Library by putting Arrays, Numpy's functions and Broadcasting to work

Develop understanding of how Arrays work and what advantages they offer over other Data Structures
Use Arrays as Data containers for common data operations
Compare time performance of your process codes versus a suitable Numpy function
In-depth understanding to use numpy's where() and select() functions to replace conventionally used methods
Apply Array Broadcasting in your line of work to replace Nested For loops and Cross-join operations

The course covers three key areas in Numpy:

Numpy Arrays as Data Structures - Developing an in-depth understanding along the lines of:
1. Intuition of Arrays as Data Containers
2. Visualizing 2D/3D and higher dimensional Arrays
3. Array Indexing and Slicing - 2D/3D Arrays
4. Performing basic/advanced operations using Numpy Arrays
Useful Numpy Functions - Basic to Advanced usage of the below Numpy functions and how they perform compared to their counterpart methods
1. numpy where() function
  1. Comparison with Apply + Lambda
  2. Performance on Large DataFrames
  3. Varied uses in new variable creation
2. numpy select() function
  1. Apply conditions on single and multiple numeric variables
  2. Apply conditions on categorical variable
Array Broadcasting - Developing an intuition of "How Arrays with dissimilar shapes interact" and how to put it to use
1. Intuition of Broadcasting concept on 2D/3D Arrays
2. Under what scenarios can we use Broadcasting to replace some of the computationally expensive methods like For loops and Cross-join Operations, etc. especially when working on a large Datasets

The course also covers the topic - "How to time your codes/processes", which will equip you to:

Track time taken by any code block (using Two different methods) and also apply to your own processes/codes
Prepare for the upcoming Chapter "Useful Numpy Functions", where we not only compare performance of Numpy functions with other conventionally used methods but also monitor how they perform on large Datasets