Daily Dose of Data Science

Home
Archive
About

NumPy

New
NumPy's methods are already highly optimized for performance. Yer, here's how you can further speed them up. Bottleneck provides a suite of optimized…
Avi Chawla
Pandas assumes that the data is a sample of the population and that the obtained result can be biased towards the sample. Thus, to generate an unbiased…
Avi Chawla
While creating conditional columns in Pandas, we tend to use the 𝐚𝐩𝐩𝐥𝐲() method almost all the time. However, 𝐚𝐩𝐩𝐥𝐲() in Pandas is nothing but…
Avi Chawla
Numpy already offers fast and optimized vectorized operations. Yet, it does not support parallelism. This provides further scope for improving the…
Avi Chawla
Matrix multiplication is a common operation in machine learning. Yet, chaining repeated multiplications using 𝐦𝐚𝐭𝐦𝐮𝐥 function makes the code…
Avi Chawla
The 𝐧𝐩.𝐦𝐚𝐭𝐦𝐮𝐥() and 𝐧𝐩.𝐝𝐨𝐭() methods produce the same output for 2D (and 1D) arrays. This makes many believe that they are the same and can…
Avi Chawla
We often print raw numpy arrays during debugging. But this approach is not very useful. This is because printing does not convey much information about…
Avi Chawla
While creating conditional columns in Pandas, we tend to use the 𝐚𝐩𝐩𝐥𝐲() method almost all the time. However, 𝐚𝐩𝐩𝐥𝐲() in Pandas is nothing but…
Avi Chawla
Polynomial linear regression using Sklearn is tedious as one has to explicitly code its features. This can get challenging when one has to iteratively…
Avi Chawla
© 2023 Avi Chawla
Privacy ∙ Terms ∙ Collection notice
Start WritingGet the app
Substack is the home for great writing