sohang3112

Machine Learning Notes

ML Concepts

• Dealing with Missing Values: For dealing with missing values in some columns of training data, one approach is to just drop the columns having missing values. But this is not recommended because the column could have important information. Instead a better approach is to fill in missing values with the column's average using sklearn.impute.SimpleImputer.

• Dealing with Categorical Data: Categorical data are like Enums - they have one of a fixed set of values (eg. male, female, other). 2 approaches:

sohang3112

# ANSI escape codes
OSC = "\033]8;;"
ST = "\033\\"
print(f"{OSC}https://example.com{ST}Example Link Text{OSC}{ST}")

sohang3112

def tee(it):
    it = iter(it)
    cache = []
    counts = [0,0]

    class _Iterator:
        def __init__(self, i, j):
           self.i = i
           self.j = j

sohang3112

from urllib.parse import urlparse, parse_qs

safe_url = 'https://apc01.safelinks.protection.outlook.com/?url=https%3A%2F%2Foneconnect.hcltech.com%2Fnav_to.do%3Furi%3Dsysapproval_approver.do%253Fsys_id%3Dca1525483b4e32101de91624c3e45a99%2526sysparm_record_target%253dsysapproval_approver%2526sysparm_stack%3Dsysapproval_approver_list.do%253Fsysparm_query%3Dsys_created_on%253ejavascript%253ags.beginningOfToday()&data=05%7C02%7Csohang.chopra%40hcltech.com%7Cd4ac502f649d4ea5e89608de407bc9c7%7C189de737c93a4f5a8b686f4ca9941912%7C0%7C0%7C639019097585714357%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=Neefy3zdm1PRH0raNr9BS3IzBpNqpt7y4EB07D43tzY%3D&reserved=0'
extracted_url = parse_qs(urlparse(safe_url).query)['url'][0]
print('Extracted URL:', extracted_url)
# Extracted URL: https://oneconnect.hcltech.com/nav_to.do?uri=sysapproval_approver.do%3Fsys_id=ca1525483b4e32101de91624c3e45a99%26sysparm_record_target%3dsysapproval_approver%

sohang3112

MathJAX NOTES

Reference: MathJAX Macros table - macros start with \. Github Markdown allows a limited subset of MathJAX and no MathJAX extensions can be used. But in static site generators, extensions can be installed using <script> tag.

Inline expressions are written within $ $, block expressions within $$ $$. Curly Brackets { } are required for grouping multiple symbols where otherwise only a single character/symbol is expected

• eg. after _ (superscript under) and ^ (power/subscript over).

sohang3112

# 'Squash Merge' video by anthonywritescode, this is script shown at timestamp 16:50 :
# https://youtu.be/5_8FTivl8Vs?si=krcdLDevrAC_DLuo&t=1013
# Uses https://github.com/newren/git-filter-repo (python script)

# Explanation: `git cat-file commit COMMIT_ID` shows parent commit id
#    normal commits have 1 parent, merge commits have 2 parents
#    this script simply deletes 1 parent (keeping only parent id of main branch), 
#    converting merge commit into normal (squash) commit

# An impressive example of this script is shown, 

sohang3112

• Generate new GPG key on laptop:

$ gpg --full-generate-key
$ gpg --list-secret-keys --keyid-format=long
$ gpg --armor --export GPG_KEY_ID    # get key id from prev list command - see details in above link

Copy output of last command (gpg armor export). Go to Github > Settings > SSH and GPG Keys > New GPG Key and paste it.

sohang3112

#! /usr/bin/env python3
import os

import fitz  # PyMuPDF
import nbformat

# Source: https://www.reddit.com/r/learnpython/comments/1ji0qwz/comment/mjblesy/?utm_source=share&utm_medium=web3x&utm_name=web3xcss&utm_term=1&utm_content=share_button
def pdf_to_ipynb(pdf_path: os.PathLike, ipynb_path: os.PathLike) -> None:
    """Convert PDF -> Jupyter Notebook .ipynb file."""
    doc = fitz.open(pdf_path)

sohang3112

{-# LANGUAGE BangPatterns #-}

import Data.List (foldl')

-- | Compresses a string by counting consecutive repeating characters.
--
-- Returns a list of tuples where each tuple consists of a character and the
-- number of its consecutive occurrences in the string.
--
-- This function uses 'reverse', 'foldl'' and bang patterns for strict accumulation. 

sohang3112

# Example of forking ziglings repo from CodeBerg to Github
UPSTREAM=https://codeberg.org/ziglings/exercises.git
FORK=git@github.com:sohang3112/ziglings.git   # mk empty Github repo & copy SSH url (as Github doesn't support push with HTTPS)

git clone $UPSTREAM ziglings/   # (optional argument) used custom cloned folder name ziglings/
cd ziglings/                    # go to cloned repo
git remote set-url origin $FORK
git remote add upstream $UPSTREAM
git push