ekzhang

" Begin vim-plug
set nocompatible
call plug#begin('~/.vim/plugged')
Plug 'VundleVim/Vundle.vim'                    " Vundle
Plug 'rstacruz/sparkup', {'rtp': 'vim/'}       " Emmet-like HTML expansion
Plug 'w0ng/vim-hybrid'                         " Hybrid theme
Plug 'sheerun/vim-polyglot'                    " Enhanced language support
Plug 'scrooloose/nerdtree'                     " Tree explorer
Plug 'vim-scripts/delimitMate.vim'             " Auto-close parens/quotes
Plug 'tpope/vim-sleuth'                        " Auto-detect indentation

ekzhang

ekzhang

Putting Big Things in Tiny Boxes: An Introduction to Compression (1h50m, ★★★)

You take a photo on your iPhone X's 12MP camera; that's 12 million pixels! This is roughly 36 MB of data -- but it only takes about 1 MB of space on your computer. How were you able to store the entire image in 1/36 of its actual size?

The magic: compression algorithms!

We will be discussing various techniques for the compression of data, starting from the basics (RLE) and ranging to lossless methods such as entropy encoding, LZW compression, and the Burrows-Wheeler Transform.

The last part of the class will be an exploration of compression methods for images, taking a look at some ad hoc approaches that we might try (8-bit graphics!). We'll then conclude with an in-depth examination of how modern lossy algorithms, like JPEG, are able to achieve ~20x compression ratios with no perceptible change in quality.

ekzhang

// Interop for sha512.ll
// Reference: http://www.iwar.org.uk/comsec/resources/cipher/sha256-384-512.pdf
// Author: Eric Zhang

#include <inttypes.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>

ekzhang

{
    "cmd": ["g++", "${file}", "-o", "${file_path}/${file_base_name}", "-std=c++14", "-O2", "-Wall"],
    "file_regex": "^(..[^:]*):([0-9]+):?([0-9]+)?:? (.*)$",
    "working_dir": "${file_path}",
    "selector": "source.c, source.c++, source.cpp",

    "variants":
    [
        {
            "name": "Run",

ekzhang

ekzhang

import copy
import random

import numpy as np
import pandas as pd
from scipy.stats import norm

pa, pb = pd.read_json('voters.json').voters_by_block
population_a, population_b = np.array(pa), np.array(pb)
population = population_a + population_b

ekzhang

ekzhang

#include <bits/stdc++.h>
using namespace std;

/* Maximum-Flow solver using Dinic's Blocking Flow Algorithm.
 * Time Complexity:
 *   - O(V^2 E) for general graphs, but in practice ~O(E^1.5)
 *   - O(V^(1/2) E) for bipartite matching
 *   - O(min(V^(2/3), E^(1/2)) E) for unit capacity graphs
 */
template<int V, class T=long long>

ekzhang

#include <bits/stdc++.h>
using namespace std;

/* Minimum-Cost, Maximum-Flow solver using Successive Shortest Paths with Dijkstra and SPFA-SLF.
 * Requirements:
 *   - No duplicate or antiparallel edges with different costs.
 *   - No negative cycles.
 * Time Complexity: O(Ef lg V) average-case, O(VE + Ef lg V) worst-case.
 */
template<int V, class T=long long>