kmdupr33

def show_doubt_if_modeling():
    global _o_predicates
    if 'model' in dc_code and 'X' in dc_code and 'y' in dc_code:
        from doubtlab.ensemble import DoubtEnsemble
        from doubtlab.reason import ProbaReason, WrongPredictionReason

        # Next we can add reasons for doubt. In this case we're saying
        # that examples deserve another look if the associated proba values
        # are low or if the model output doesn't match the associated label.
        reasons = {

kmdupr33

Prediction summary	Date	Confidence	Came true?	Prediction Squared Error	Brier Score
Finish programming task X	10/31/20	.85	TRUE	0.0625	0.0625
Refactoring will facilitate product work	1/31/20	.75	FALSE	0.5625	0.3125

kmdupr33

Tasks	area
Implement rename project	projects
Implement google login	accounts
Implement archive project	projects

kmdupr33

Files	Changes for
login.js	accounts
project.js	projects

kmdupr33

/** changes-for: accounts */

function login() {
  /* */
}

function isLoggedIn() {
  /* */
}

kmdupr33

#!/usr/bin/env bash
set -euo pipefail

file="$1"

last_sha="$(git rev-list HEAD | tail -n 1)"

function sum_numbers() {
  perl -lane '$t+=$_ for @F; print $t'
}

kmdupr33

fun updateGuess(
    guess: Pair<Double, Double>,
    learningRate: Double,
    videoGameData: Array<Pair<Double, Double>>
): Pair<Double, Double> {
    val (m, b)  = guess
    val mGradient = videoGameData.fold(0.0) { acc: Double, pair: Pair<Double, Double> ->
        val (x, y) = pair
        acc + ((((m * x) + b) - y) * x)
    } / videoGameData.size

kmdupr33

fun cost(videoGameData: Array<Pair<Double, Double>>, learningParameters: Pair<Double, Double>): Double =
    videoGameData.fold(0.0) { acc: Double, pair: Pair<Double, Double> ->
        val (m, b) = learningParameters
        val (x, y) = pair
        acc + ((((m * x) + b) - y).pow(2)/2)
    } / videoGameData.size

kmdupr33

fun findLearningParameters(videoGameData: Array<Pair<Double, Double>>): Pair<Double, Double> {
    val learningRate = .0003
    var guess = Pair(0.0, 0.0)
    for (i in 1..10000000) {
        guess = updateGuess(guess, learningRate, videoGameData)
        if (i % 1000 == 0) println("Guess: $guess")
        if (i % 1000 == 0) println("Cost: ${cost(videoGameData, guess)}")
    }
    return guess
}

kmdupr33

fun predictVideoGameSales(
  metaCriticScore: Int,
  learningParameters: Pair<Double, Double>
) {
  val (m, b) = learningParameters
  return metaCriticScore * m + b
}