piyush01123

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

int randGen (int i) {srand(time(0)); return rand()%i;}

int partition(int A[], int lo, int hi)
{
    int pivot = A[hi];
    int i = lo-1;
    for (int j=lo; j<=hi; j++)

piyush01123

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

int randGen (int i) {srand(time(0)); return rand()%i;}
int ctr = 0;

int partition(int A[], int lo, int hi)
{
	int pivot = A[hi];
	int i = lo-1;

piyush01123

// ==UserScript==
// @name         Auto Wayback
// @namespace    http://tampermonkey.net/
// @version      3.1.1
// @description  Load latest snapshot from Wayback Machine
// @author       Piyush Singh
// @match        *://*/*
// @icon         data:image/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==
// @grant        unsafeWindow
// ==/UserScript==

piyush01123

Anomaly detection for a large industrial facility

About: This document contains summary of a real-time anomaly detection system developed by me (and team) for Anadarko's crude oil extraction facilities during my time at Quantiphi.

Requirements

• Our client has facilities of 2 types. For both, we want to have models that can do anomaly detection in real time.
• For type I we have some samples marked as anomalies, so we can either train a supervised ML model (with the challenge of an extreme bias dataset) or we can use the model trained for type II dataset but with some finetuning using the annotation. We go with the 2nd route.
• For type II we do not have any marked samples and we have to train a model in unsupervised setting
• We want high recall with decent precision since false negatives incur higher costs (replacing entire components) whereas false positives incur marginal costs (extra component checks)

piyush01123

// ==UserScript==
// @name         Generate Strong Password
// @namespace    http://tampermonkey.net/
// @version      3.1.1
// @description  Strong password generator of size 12
// @author       Piyush Singh
// @match        *://*/*
// @icon         data:image/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==
// @grant        unsafeWindow
// @grant        GM_setClipboard

piyush01123

Command Line Productivity Tools

In course of my everyday work, sometimes I need a tool which does not exist. So I build them myself and later release them. Here are 3 command line tools I have built that each serve some specific purpose:

Tool 1: Flask Image Gallery

Generates image gallery from your server. Very useful during ML experimentation. Link to repository.

piyush01123

from flask import Flask, request
from datetime import datetime
import json,os

app = Flask(__name__)
@app.route('/', methods=["GET","POST"])
def index():
	if request.method=="POST":
		ip_data = request.get_json()
		print(ip_data)

piyush01123

float fisr(float n){
	long x = 0x5f3759df - (*(long *)&n >> 1); //magic number = 1.5*2^23*(127-mu); mu represents log(1+x)=x+mu; mu=0.0450465
	float y = *(float *)&x;  // Solution from bit manipulation
	y = 1.5F*y - 0.5F*n*y*y*y; // Newton improvement 1
        y = 1.5F*y - 0.5F*n*y*y*y; // Newton improvement 2
	return y;
}

piyush01123

import numpy as np

def viterbi(mood_sequence, priors, transmission_probs, emission_probs):
    n = len(mood_sequence)
    weather_matrix = np.zeros((n, 2))
    history = [(None,None)]
    for i, mood in enumerate(mood_sequence):
        if i==0:
            weather_matrix[i] = priors['s']*emission_probs['s'+mood], priors['r']*emission_probs['r'+mood]
        else:

piyush01123

from sklearn.feature_extraction.text import TfidfTransformer
import numpy as np

counts = [[3, 0, 1],
          [2, 0, 0],
          [3, 0, 0],
          [4, 0, 0],
          [3, 2, 0],
          [3, 0, 2]]