Yankim

import os
from PIL import Image
import pathlib
PATH_ = pathlib.Path(__file__).parent
broken_images=[]
def check_images(PATH_):   
    for pic_class in os.listdir(PATH_):
        for pic in os.listdir(f'{PATH_}/{pic_class}'):
            try:
                img = Image.open(f'{PATH_}/{pic_class}/{pic}') 

Yankim

#Generate County map with a few var of interest
percent_map <- function(var, color, legend.title, min = 0, max = 100, name = "") {
    # generate vector of fill colors for map
    shades <- colorRampPalette(c("white", color))(100)
    # constrain gradient to percents that occur between min and max
    var <- pmax(var, min)
    var <- pmin(var, max)
    percents <- as.integer(cut(var, 100,
                               include.lowest = TRUE, ordered = TRUE))
    fills <- shades[percents]

Yankim

#Stepwise regression was used to predict obesity rate. Starting from saturated model
#variables not found significant were removed using stepwise regression.
data = fulldb[,-c(1:3)]
complete.data = data[complete.cases(data),]
model.saturated = lm(PCT_OBESE_ADULTS10 ~ ., data = complete.data)
model.empty = lm(PCT_OBESE_ADULTS10 ~ 1, data = complete.data)
scope = list(lower = formula(model.empty), upper = formula(model.saturated))
backwardAIC = step(model.saturated, scope, direction = "backward", k = 2)

#Used to predict obesity rate with multiple linear regression

Yankim

#Generate County map with a few var of interest
percent_map <- function(var, color, legend.title, min = 0, max = 100, name = "") {
    # generate vector of fill colors for map
    shades <- colorRampPalette(c("white", color))(100)
    # constrain gradient to percents that occur between min and max
    var <- pmax(var, min)
    var <- pmin(var, max)
    percents <- as.integer(cut(var, 100,
                               include.lowest = TRUE, ordered = TRUE))
    fills <- shades[percents]

Yankim

#Predicting the test data
xgmat.test <- xgb.DMatrix(as.matrix(higgs.test.dummy))
xgboostTestPred <- predict(bst, newdata=xgmat.test)
  
predicted <- rep("s",550000)
predicted[xgboostTestPred>=threshold] <- "b"
weightRank = rank(xgboostTestPred, ties.method= "random")

write.csv(as.data.frame(xgboostTestPred), "Submissions/xgboost_prob.csv")
write.csv(as.data.frame(higgs.testId), "Submissions/EventID.csv")

Yankim

#Predicting training data
xgmat.train <- xgb.DMatrix(as.matrix(higgs.train.dummy), 
                           label = as.numeric(higgs.labels == "X0"),
                           weight = scaled.weight)
xgboostTrainPred <- predict(bst, newdata = xgmat.train)
labels <- ifelse(as.character(higgs.labels)=="X1", 1, 0)

auc = roc(labels, xgboostTrainPred)
plot(auc, print.thres=TRUE)
######## From the graph, we can tell the best threshold is 0.002

Yankim

# Grid for the parameter search
#The guidlines for how to tune parameters are commented below and are taken from
# Owen Zheng http://www.slideshare.net/OwenZhang2/tips-for-data-science-competitions
xgb_grid_1 = expand.grid(
  eta = c(.5, 1, 1.5),                #[2-10]/num trees
  max_depth = c(4, 6, 8),             #Start with 6
  nrounds = 100,                      #Fix at 100
  gamma = 0,                          #Usually ok to leave at 0
  colsample_bytree = c(.3, .5, .7),   #.3 - .5
  min_child_weight = 1                #start with 1/sqrt(eventrate)

Yankim

library(xgboost); library(methods); library(pROC); library(caret); library(xgboost); library(readr); library(plyr); library(dplyr)
library(tidyr); library(dummies); library(doMC); registerDoMC(cores = 4)

#Read in the data
#higgs.___.full is raw data
higgs.train.full = read.csv('./data/training.csv', header=T)
higgs.test.full = read.csv('./data/test.csv', header=T)
higgs.testId = higgs.test.full$EventId

#############################################

Yankim

library(shiny); library(shinydashboard); library(plotly); source("helpers.R"); library(DT)

shinyUI(dashboardPage(
    dashboardHeader(title = "Food and health demographics in the USA", titleWidth = 400), 
    dashboardSidebar(
        sidebarUserPanel("Yannick Kimmel", image = "Yannick.jpg"),
        
        sidebarMenu(
            menuItem("Map", tabName = "mappanel", icon = icon("map")),
            menuItem("Trends", tabName = "trends", icon = icon("line-chart")),

Yankim

# server.R
library(shiny); library(maps); library(mapproj); library(shinydashboard); library(plotly); library(DT); source("helpers.R")

shinyServer(
    function(input, output) {
        output$map <- renderPlot({
            args <- switch(input$var,
                           "Percent Adult Obese 2009" = list(health2$PCT_OBESE_ADULTS09, "darkgreen", "% Obese"),
                           "Percent Adult Obese 2010" = list(health2$PCT_OBESE_ADULTS10, "darkgreen", "% Obese"),
                           "Percent Adult Diabetic 2009" = list(health2$PCT_DIABETES_ADULTS09, "darkred", "% Diabetic"),