Chapter 8 Data visualization Team
8.1 Biometric Wearables
This project is a part of the Purdue data science initiative, where students are given an opportunity to work on the data-driven project, guided by an industry mentor. This Biometric Wearables project is a part of an ongoing collaboration between the Purdue and Merck industry. Following is the brief description of the project:
8.2 Purpose
The purpose of the project is to create an automated system that will capture the biometric data from wearable fitness technology using the mobile application and communicate its’ result to Merck scientists.
8.3 Outcome
This automated system will be able to collect accurate and less biased data from the clinical trial patients because the digital process reduces the human error and reduce the pressure on the clinical trial patient with manual processing procedure.
8.4 Significance
This project may help to reduce the experiment biased by collecting more extensive data and using the technology. Also, this approach may help to reduce the clinical trial time and cost to ease of data collection.
8.5 Overview of Data visualization Team
This team is a part of a big team working on the above-described project, part of the Merck-Purdue data science collaboration. This team has been assigned the responsibility to design rich, interactive graphics, data visualization, and plotting method to facilitate the MERCK scientists. In addition, this team is responsible for brainstorming the data, and experiments with different visualization techniques that can help the project.
8.6 Previous Work Done on the project
The previous team that worked on the data visualization aspect of this project created a baseline shiny website that had some specific widgets created. There are two pages on the shiny website i.e., activity progress and step goals. The activity progress has a bar chart that shows the progress by comparing two biometric data points. The bar chart has an x-axis which is to show the dates and a y-axis which is to show the steps. One of the widget’s features on this page was created to view a specific time period of the data. The second page is to show step goals using a pie chart.
8.7 Team Members
Following is the brief introduction of the data visualization team members: (Ahmed Ashraf Butt and Sharon Patta participated in the first semester)
8.7.1 Ahmed Ashraf Butt
He is a doctoral student at the School of Engineering Education, Purdue University. He is currently working as a research assistant on the CourseMIRROR project funded by the Institute of Education Sciences (IES). He is interested in designing educational tools and exploring their impact on enhancing students’ learning experiences. Also, he is working on the Merck-Purdue data science collaboration responsible for creating an automated system that will capture the biometric data from the wearable fitness technology using the mobile application and communicate its result to Merck scientists.
8.7.2 Pika Seth
She is a second-year student in the College of Engineering studying Biomedical Engineering at Purdue University. She is currently involved with the Data Mine Corporate Program working on the Biometrics Wearable Project with Merck. She is involved with the data visualization aspect dealing with the programming of R Shiny. She is a part of the Red Cross Club and the Women in Engineering Program. Also, she is a part of the Sigma Delta Tau sorority.
8.7.4 Surya Suresh
He is a first-year student in the College of Science at Purdue University studying Computer Science. He is working with the Merck Data Mine group on the Data Visualization part of the project. He is responsible for creating new visualisations to add to the R Shiny Dashboard.
8.8 Team formation
The team has worked together to become more familiar with R programming and shiny application. Each member of the group explored different data visualizations and then, worked colloborately to implement a new visualization or improving the previous shiny app. Additionally, the group had several meetings to collectively present findings to one another and discuss errors and troubleshooting.
8.9 Term 1 goals
We are going to begin working on adding more features to the old Shiny App as well as implementing our own visualizations that we think would be useful. We also need to figure out where we are getting our data from once the API is running because as of right now we are just taking it off of the CSV file that has Dr. Ward’s data.
8.10 Visulization 1
In this visualization, we were able to create a visualization that adapts to the given data and creates a Data Table with all of the given data. We used the Data Tables library to accomplish this. The Data Table contains two main tabs. One tab contains all of the mean values of the given data. The other tab allows the user to navigate through all of the data points in a clean and efficient way.Specifically it allows the user to filter the data as well as list the data in any desired order. For example, the data can be listed in order from the days with the most steps to the days with the least.
8.10.1 Semester 1 Code
I imported the libraries needed for the visualization
library(shiny) #Used for Shiny Dashboard
library(ggplot2) #Needed for Plots (Not specifically this Viz)
library(reshape2) #Needed for Plots (Not specifically this Viz)
library(DT) #Used for the Data Tables used in this Vizualizationbiometric data 'bioFinal.csv' for biometric data from 2020 Feb. to present? - Merckteam's data 'DrWardData.csv' for biometric data from 2016 to 2019 April - Dr. Ward's data
Used to read data into the myDF data frame
myDF <- read.csv("liveMerck.csv",TRUE,",")active miles (stacked)
miles1 <- subset(myDF, select = c('Date','Lightly.Active.Miles',
'Moderately.Active.Miles',
'Very.Active.Miles'))
miles <- melt(miles1)
miles$Date <- as.Date(miles$Date)active minutes (stacked)
minutes1 <- subset(myDF, select = c('Date', 'Sedentary.Minutes',
'Lightly.Active.Minutes',
'Fairly.Active.Minutes',
'Very.Active.Minutes'))
minutes <- melt(minutes1)
minutes$Date <- as.Date(minutes$Date)HR minutes (stacked)
hr1 <- subset(myDF, select = c('Date', 'HR.30.100.Minutes',
'HR.100.140.Minutes', 'HR.140.170.Minutes',
'HR.170.220.Minutes'))
hr <- melt(hr1)
hr$Date <- as.Date(hr$Date)Finding out the collums index with numeric value in it.
columsIndexWithNumberValue <- unlist(lapply(myDF, is.numeric)) Extracting the columns with number value in it.
columWithNumber <- myDF[ , columsIndexWithNumberValue]finding the mean of all collum's mean
meansOFColumWithNumber <- lapply(columWithNumber[,], mean,na.rm=TRUE)Converting them back into data frame.
meansOFColumWithNumberDataFrame<-data.frame(matrix(unlist(meansOFColumWithNumber), nrow=length(meansOFColumWithNumber), byrow=T))Extracting the name of the collums names from columWithNumber.
colName<- data.frame(names(columWithNumber))Extracting the name of the collums.
finalDataFrame <- data.frame(a=colName, c=meansOFColumWithNumberDataFrame)changing the Collum name of the final dataframe
colnames(finalDataFrame)[1]<-"Variables"
colnames(finalDataFrame)[2]<-"Mean Values"myDF to date format
myDF$Date = as.Date(myDF$Date)Remove the X column from myDF
drops <- c("X")
myDF <- myDF[ , !(names(myDF) %in% drops)]UI of the Application
# Uses the fluidPage and a Sidebar Layout
ui <- fluidPage(
title = "Tester",
titlePanel("Visualization 1"),
sidebarLayout(
sidebarPanel(
# This panel contains the first Data Table
conditionalPanel(
#input.dataset is used in DT's to show which data is actually used
'input.dataset === "myDF"',
#Included a checkbox for what columns should be in the data Table
#The pre selected columns are "Date" and "Steps"
checkboxGroupInput("show_vars", "Columns of information to show:",
names(myDF), selected = list("Date", "Steps"))
),
# This panel contains the second Data Table
conditionalPanel(
#input.dataset is used in DT's to show which data is actually used
'input.dataset === "finalDataFrame"',
#We can change this text to anything we want user to see
helpText("Click the column header to sort a column.")
),
),
mainPanel(
# Allows the user to pick between the 2 tabs containing the Data Tables
tabsetPanel(
#name of the tabs
id = 'dataset',
#First tab named "myDF" contains Data Table from first set of data
tabPanel("myDF", DT::dataTableOutput("mytable1")),
#Second tab named "finalDataFrame" contains DT from second set of data
tabPanel("finalDataFrame", DT::dataTableOutput("mytable2"))
)
)
)
)Server of the Application
server <- function(input, output) {
# choose columns to display
myDF2 = myDF[sample(nrow(myDF), nrow(myDF), replace = FALSE), ]
#Used to render the actual Data Table based off any changes made by the user
#Changes based on what variables are selected and the filter feature is
#located on the top of the Data Table
output$mytable1 <- DT::renderDataTable({
DT::datatable(myDF2[, input$show_vars, drop = FALSE], rownames = FALSE, filter = 'top')
})
#Used to render teh Data Table and does not have many features like the first
#Data Table. This one only contains the means (finalDataFrame)
output$mytable2 <- DT::renderDataTable({
DT::datatable(finalDataFrame, options = list(orderClasses = TRUE), rownames = FALSE)
})
}Runs the application
shinyApp(ui, server)8.11 Visulization 2
In this visulization, we categorizes the data into three categories: Steps and Floors Climbed, Distance, and Activity Level. Each category contains bar graphs that display the data for each respective variable per month. Both the Distance and the Activity Level categories display visualizations that separate the data into different subcategories based on the intensity of the activity.
8.11.1 Code
library(shiny)
library(ggplot2)
library(reshape2)
# biometric data
# 'bioFinal.csv' for biometric data from 2020 Feb. to present? - Merckteam's data
# 'DrWardData.csv' for biometric data from 2016 to 2019 April - Dr. Ward's data
#This function reads in the CSV file that contains the data, named DrWardData.csv, and stores it in a data frame called myDF. The argument for the read.csv function is the path where the CSV file is located.
myDF <- read.csv("/class/datamine/corporate/merck/Merck201920/fitbit_data/DrWardData.csv")
# The subset function selects the data from the myDF data frame that separates the miles the user has walked by activity level, as well as the date column. This subset is stored as "miles1".
# The melt function then takes the columns stored in miles1 and stacks them into a single column in "miles".
# The as.Date function goes into the Date column of "miles" and correctly formats the values as a date.
miles1 <- subset(myDF, select = c('Date','Lightly.Active.Miles', 'Moderately.Active.Miles', 'Very.Active.Miles'))
miles <- melt(miles1)
miles$Date <- as.Date(miles$Date)
# The subset function selects the data from the myDF data frame that separates the minutes the user spends in activity by activity level, as well as the date column. This subset is stored as "minutes1".
# The melt function then takes the columns stored in minutes1 and stacks them into a single column in "minutes".
# The as.Date function goes into the Date column of "minutes" and correctly formats the values as a date.
minutes1 <- subset(myDF, select = c('Date', 'Sedentary.Minutes', 'Lightly.Active.Minutes', 'Fairly.Active.Minutes', 'Very.Active.Minutes'))
minutes <- melt(minutes1)
minutes$Date <- as.Date(minutes$Date)
# The subset function selects the data from the myDF data frame that separates the data by the number of minutes the user spent with heart rate between 30-100 bpm, 100-140 bpm, 140-170 bpm, and 170-220 bpm, as well as the date column. This subset is stored as "hr1".
# The melt function then takes the columns stored in hr1 and stacks them into a single column in "hr".
# The as.Date function goes into the Date column of "hr" and correctly formats the values as a date.
hr1 <- subset(myDF, select = c('Date', 'HR.30.100.Minutes', 'HR.100.140.Minutes', 'HR.140.170.Minutes', 'HR.170.220.Minutes'))
hr <- melt(hr1)
hr$Date <- as.Date(hr$Date)
# The as.Date function goes into the Date column of "myDF" and correctly formats the values as a date.
# The first format function goes into the year column of "myDF" and formats the values as a year.
# The second format function goes into the month column of "myDF" and formats the values as unabbreviated month names.
myDF$Date = as.Date(myDF$Date)
myDF$year <- format(myDF$Date,'%Y')
myDF$month <- factor(format(myDF$Date,'%B'), levels = month.name)
# Every shiny app must include a UI header beginning with fluidPage()
# fluidPage() allows the user to adjust the size of their window while preserving the dimensions of the display
ui<-fluidPage(
# The first section of code uses the navbarPage() function to create the menu at the top of the page titled “Fitbit Dashboard” that allows the user to select the “How to Use” page, the “Activity Progress” page, or the “Step Goals” page
# Creates a menu of selections titled "Fitbit Dashboard"
navbarPage("Fitbit Dashboard",
# tabPanel 1
# The tabPanel() function creates the “How to Use” tab in the Fitbit Dashboard and includes the images, displays, and texts to be shown on the page
tabPanel("How to Use",
# Includes Merck logo
img(src = "https://assets.phenompeople.com/CareerConnectResources/MERCUS/social/1200x630-1552902977090.jpg", height = 120, width = 220),
# Creates main title "Biometric Data Dashboard"
h1("Biometric Data Dashboard"),
mainPanel(
# Includes bolded text and link
strong("Refer to the website here for more information:"),
a("https://www.merck.com/index.html"),
# Smaller heading than h1 to create title "How to Use My Dashboard"
h3("How to Use My Dashboard"),
# Outputs HTML "text2" which is included in the server block
htmlOutput("text2")
)
),
# tabPanel 2
# The tabPanel() function creates the “Activity Progress” tab in the Fitbit Dashboard and includes the images, displays, and texts to be shown on the page
tabPanel("Activity Progress",
sidebarLayout(
# Creates title "Activity Progress" for sidebar panel
sidebarPanel(h3("Activity Progress"),
# Allows user to input date range
dateRangeInput("dates", ("Date range")),
# Allows user to select single y-axis variable for graph
varSelectInput("variables", ("Variable:"), myDF[c(-1, -2)]),
# Allows user to select variable y-axis (includes all levels of selected on the same graph)
selectInput("stack", ("Variable (Stacked):"),
c("N/A" = "nana", "Active Miles" = "miles", "Active minutes" = "minutes", "HR minutes" = "hr")),
),
# Creates the main panel
mainPanel(
# Includes italicized text
em("Select a date range and a variable!"),
# Includes bolded text
strong("Refer to the website here for more information:"),
# Includes link
a("https://www.merck.com/index.html"),
# Displays the bar graph with selected variable
plotOutput(outputId = "bioBarGraph")
)
)
),
# tabPanel 3
# The tabPanel() function creates the “Step Goals” tab in the Fitbit Dashboard and includes the images, displays, and texts to be shown on the page
tabPanel("Step Goals",
# Creates sidebar panel with smaller heading "10,000-Step Goal"
sidebarPanel(h3("10,000-Step Goal"),
# Allows user to input one specific date
dateInput("goaldate", ("Date"))
),
# Creates the main panel
mainPanel(
# Includes bolded text
strong("Refer to the website here for more information:"),
# Includes link
a("https://www.merck.com/index.html"),
# Displays pie chart
plotOutput(outputId = "bioDoughnut")
)
),
# tabPanel 4
# The tabPanel() function creates the “Visualization 2” tab in the Fitbit Dashboard and includes the images, displays, and texts to be shown on the page
tabPanel("Visualization 2",
sidebarLayout(
# Creates title "Activity Progress" for sidebar panel
sidebarPanel(h3("Visualization 2"),
),
# Creates the main panel
mainPanel(
# Includes bolded text
strong(HTML("Activity graphs are listed below <br/>")),
strong("Steps and Floors Climbed"),
# Displays the bar graph that shows total steps per month
plotOutput(outputId = "bioStep"),
# Displays the bar graph that shows total floors climbed per month
plotOutput(outputId = "bioFloors"),
# Includes bolded text
strong(HTML("Distance <br/>")),
# Includes italicized text
em("This section shows the total distance walked per month in miles, as well as a breakdown of the activity level of the distance walked."),
# Displays the bar graph that shows total miles walked per month
plotOutput(outputId = "bioTM"),
# Displays the bar graph that shows total number of lightly active miles walked per month
plotOutput(outputId = "bioLM"),
# Displays the bar graph that shows total number of moderately active miles walked per month
plotOutput(outputId = "bioMM"),
# Displays the bar graph that shows total number of very active miles walked per month
plotOutput(outputId = "bioVM"),
# Includes bolded text
strong(HTML("Activity Level <br/>")),
# Includes italicized text
em("This section shows the time spent in each level of activity."),
# Displays the bar graph that shows the amount of time spent being lightly active per month
plotOutput(outputId = "bioLA"),
# Displays the bar graph that shows the amount of time spent being fairly active per month
plotOutput(outputId = "bioFA"),
# Displays the bar graph that shows the amount of time spent being very active per month
plotOutput(outputId = "bioVA"),
)
)
)
)
)
server <- function(input, output) {
# You can access the values of the widget (as a vector of Dates)
# with input$dates,
# e.g. output$value <- renderPrint({ input$dates })
# text on tabPanel 1
output$text2 <- renderUI({
HTML(paste("", "1. Activity Progress with Bar Charts",
"- Date Range: select a start and an end date to set a date range of the activity progress",
"- Variables: use the drop-down to select a variable of the activity progress you wish to see",
"- Variables (Stacked): use the drop-down to select a variable of the activity progress you wish to see as a stacked bar chart",
"", "2. Step Goals with Doughnut Charts",
"- Date: select a date of the activity progress you wish to see",
sep="<br/>"))
})
# bar charts on tabPanel 2
output$bioBarGraph <- renderPlot({
# counter for the start date - date range
counter1 = 0
for (myDate in as.character(myDF$Date)){
counter1 = counter1 + 1
if (myDate == input$dates[1]) {
break
}
}
# counter for the end date - date range
counter2 = 0
for (myDate in as.character(myDF$Date)){
counter2 = counter2 + 1
if (myDate == input$dates[2]) {
break
}
}
# subset of myDF with the selected date range
mySubset <- myDF[c(counter1:counter2),]
if (input$stack == "nana"){ # single bar chart - if stacked bar is N/A
ggplot(data=mySubset, aes(x=Date, y=!!input$variables)) + geom_bar(stat="identity", width = .5, fill = "#13A999")
}else if(input$stack == "miles"){ # stacked bar chart - active miles
ggplot(data=miles, aes(x=Date, y=value, fill=variable)) + geom_bar(stat="identity", width = .5)
}else if(input$stack == "minutes"){ # stacked bar chart - active minutes
ggplot(data=minutes, aes(x=Date, y=value, fill=variable)) + geom_bar(stat="identity", width = .5)
}else if(input$stack == "hr"){ # stacked bar chart - HR minutes
ggplot(data=hr, aes(x=Date, y=value, fill=variable)) + geom_bar(stat="identity", width = .5)
}
})
# Visualization 2
# Bar graph that shows total steps per month
output$bioStep <- renderPlot({
ggplot(data = myDF, aes(x = month, y= Steps))+
geom_bar(stat = "identity", color = "#007a73", width = 0.5, position = position_dodge(width = 0.6), fill = "#007a73") +
ggtitle("Total Steps per Month") +
xlab("Month") +
ylab("Steps")
})
# Bar graph that shows total floors climbed per month
output$bioFloors <- renderPlot({
ggplot(data = myDF, aes(x = month, y= Floors.Climbed))+
geom_bar(stat = "identity", color = "#007a73", width = 0.5, position = position_dodge(width = 0.6), fill = "#007a73") +
ggtitle("Total Floors Climbed per Month") +
xlab("Month") +
ylab("Floors Climbed")
})
# Bar graph that shows total miles walked per month
output$bioTM <- renderPlot({
ggplot(data = myDF, aes(x = month, y= Total.Miles))+
geom_bar(stat = "identity", color = "#007a73", width = 0.5, position = position_dodge(width = 0.6), fill = "#007a73") +
ggtitle("Total Miles Walked per Month") +
xlab("Month") +
ylab("Miles")
})
# Bar graph that shows lightly active miles walked per month
output$bioLM <- renderPlot({
ggplot(data = myDF, aes(x = month, y= Lightly.Active.Miles))+
geom_bar(stat = "identity", color = "#007a73", width = 0.5, position = position_dodge(width = 0.6), fill = "#007a73") +
ggtitle("Lightly Active Miles Walked per Month") +
xlab("Month") +
ylab("Miles")
})
# Bar graph that shows moderately active miles walked per month
output$bioMM <- renderPlot({
ggplot(data = myDF, aes(x = month, y= Moderately.Active.Miles))+
geom_bar(stat = "identity", color = "#007a73", width = 0.5, position = position_dodge(width = 0.6), fill = "#007a73") +
ggtitle("Moderately Active Miles Walked per Month") +
xlab("Month") +
ylab("Miles")
})
# Bar graph that shows very active miles walked per month
output$bioVM <- renderPlot({
ggplot(data = myDF, aes(x = month, y= Very.Active.Miles))+
geom_bar(stat = "identity", color = "#007a73", width = 0.5, position = position_dodge(width = 0.6), fill = "#007a73") +
ggtitle("Very Active Miles Walked per Month") +
xlab("Month") +
ylab("Miles")
})
# Bar graph that shows the amount of time spent being lightly active per month
output$bioLA <- renderPlot({
ggplot(data = myDF, aes(x = month, y= Lightly.Active.Minutes))+
geom_bar(stat = "identity", color = "#007a73", width = 0.5, position = position_dodge(width = 0.6), fill = "#007a73") +
ggtitle("Time Spent Being Lightly Active per Month") +
xlab("Month") +
ylab("Minutes")
})
# Bar graph that shows the amount of time spent being fairly active per month
output$bioFA <- renderPlot({
ggplot(data = myDF, aes(x = month, y= Fairly.Active.Minutes))+
geom_bar(stat = "identity", color = "#007a73", width = 0.5, position = position_dodge(width = 0.6), fill = "#007a73") +
ggtitle("Time Spent Being Fairly Active per Month") +
xlab("Month") +
ylab("Minutes")
})
# Bar graph that shows the amount of time spent being very active per month
output$bioVA <- renderPlot({
ggplot(data = myDF, aes(x = month, y= Very.Active.Minutes))+
geom_bar(stat = "identity", color = "#007a73", width = 0.5, position = position_dodge(width = 0.6), fill = "#007a73") +
ggtitle("Time Spent Being Very Active per Month") +
xlab("Month") +
ylab("Minutes")
})
# doughnut chart on tabPanel 3
output$bioDoughnut <- renderPlot({
# counter for the date - daily goals
counter3 = 0
for (myDate in as.character(myDF$Date)){
counter3 = counter3 + 1
if (myDate == input$goaldate) {
break
}
}
# doughnut bar
# daily steps of the selected date
steps <- myDF[c(counter3:counter3),c("Steps")]
# get a percentage - daily steps out of 10000 steps
data <- data.frame(
category=c("Steps","N/A"),
count=c(steps, 10000-steps)
)
data$fraction = data$count / 100
data$ymax = cumsum(data$fraction)
data$ymin = c(0, head(data$ymax, n=-1))
ggplot(data, aes(ymax=ymax, ymin=ymin, xmax=4, xmin=3, fill=category)) + geom_rect() + coord_polar(theta="y") + xlim(c(2, 4))
})
}
shinyApp(ui, server)8.12 Term 2 goals
With Surya and Pika returning to the Data Mine for the spring semester, we are going to connect to the database and continue to refine visualizations that were created in the first semester of the school year. We will also be focusing on making the dashboard look more appealing with themes and adding better descriptions to our visualizations. We are also going to split the dashboard into two separate dashboards for single patient data and total FitBit data. We also need to add more visualizations to our dashboard.
I imported the libraries needed for the visualizations
library(RMariaDB) #Used to connect to the database
library(shiny) #Used for Shiny Dashboard
library(ggplot2) #Needed for Plots
library(reshape2) #Needed for Plots
library(highcharter) #Needed for averages visualization
library(dplyr) #Needed alongside Highcharter to create Averages visualization
library(shinythemes) #Needed to add theme to the visualization
library(DT) #Used for the Data Tables used in this Vizualization8.13 Activity Progress Visualisation
This visualization allows users to view all of the data as a bar plot. There is an option to choose the desired date range to view the data from as well as options to choose whichever of the variables of each data point is desired. Users are also able to view a stacked bar plot for all of the data points of different active minutes and active miles. This visualization makes it much easier for users to notice a trend over the course of the date range.
8.14 Average Statistics Visualization
This visualization allows users to view the average statistics of many different variables for each month of the year as well as each day of the week in a bar plot. Both are presented as separate visualizations but are on the same panel of the dashboard. This visualizations is useful to find trends in for example, steps taken, between different months of the year and different days of the week.
8.15 Single Patient Data Dashboard
This dashboard is to be used with only single patient data. Once we figure out how to get the desired patient from the website, we will be able to get the data points from the specific user from the database. This dashboard contains the Activity Progress Visualization, the Data Table Visualization, and the Average Statistics Visualization. This dashboard includes more patient specific data such as BMI unlike the All patients dashboard.
8.16 Combined Patient Data Dashboard
This dashboard is meant to be used with combined data of all of the patients in the database. This dashboard does not include as much data due to the fact that some data such as weight is not appropriate to view in this context. This dashboard contains the Data Table Visualization and the Average Statistics Visualization. The Activity Progress Visualization is not included because it does not make sense when using more then one patient.
8.16.1 Single Patient Visualization Code
#SINGLE PATIENT
library(RMariaDB)
library(shiny)
library(ggplot2)
library(reshape2)
library(highcharter)
library(dplyr)
library(shinythemes)
library(DT)
#Connect to the database
db <- dbConnect(RMariaDB::MariaDB(),
host = "scholar-db.rcac.purdue.edu",
db = "merck",
user = "merck_user",
password = "Ph$rma_user")
#get fitbit data from the database
myDF <- dbGetQuery(db, "SELECT * FROM fitbit_data;")
#Disconnect to database
dbDisconnect(db)
#correctly format the dates
myDF$collection_date <- sapply(myDF$collection_date, substring, 0, 10)
# active miles (stacked)
miles1 <- subset(myDF, select = c('collection_date','lightly_active_miles', 'moderately_active_miles', 'very_active_miles'))
miles <- melt(miles1)
miles$collection_date <- as.Date(miles$collection_date)
# active minutes (stacked)
minutes1 <- subset(myDF, select = c('collection_date', 'sedentary_minutes', 'lightly_active_minutes', 'fairly_active_minutes', 'very_active_minutes'))
minutes <- melt(minutes1)
minutes$collection_date <- as.Date(minutes$collection_date)
# HR minutes (stacked)
hr1 <- subset(myDF, select = c('collection_date', 'hr30_100_minutes', 'hr100_140_minutes', 'hr140_170_minutes', 'hr170_220_minutes'))
hr <- melt(hr1)
hr$collection_date <- as.Date(hr$collection_date)
# myDF to date format
myDF$collection_date = as.Date(myDF$collection_date)
myDF$year <- format(myDF$collection_date,'%Y')
myDF$month <- factor(format(myDF$collection_date,'%B'), levels = month.name)
myDF$day_of_week <- weekdays(myDF$collection_date)
#create the averages data frame that contains average statistics per month
averagesDF <- data.frame(
Var = c("Steps", "Floors Climbed", "Total Miles", "Lightly Active Miles", "Moderately Active Miles",
"Very Active Miles", "Sedentary Minutes", "Lightly Active Minutes", "Fairly Active Minutes",
"Very Active Minutes", "HR 30-100 Minutes", "HR 100-140 Minutes", "HR 140-170 Minutes",
"HR 170-220 Minutes", "Average Resting HR", "BMI", "Minutes Asleep", "Sleep Efficiency", "Weight"),
Jan = c(mean(myDF$steps[myDF$month == "January"]), mean(myDF$floors_climbed[myDF$month == "January"]),
mean(myDF$total_miles[myDF$month == "January"]), mean(myDF$lightly_active_miles[myDF$month == "January"]),
mean(myDF$moderately_active_miles[myDF$month == "January"]), mean(myDF$very_active_miles[myDF$month == "January"]),
mean(myDF$sedentary_minutes[myDF$month == "January"]), mean(myDF$lightly_active_minutes[myDF$month == "January"]),
mean(myDF$fairly_active_minutes[myDF$month == "January"]), mean(myDF$very_active_minutes[myDF$month == "January"]),
mean(myDF$hr30_100_minutes[myDF$month == "January"]), mean(myDF$hr100_140_minutes[myDF$month == "January"]),
mean(myDF$hr140_170_minutes[myDF$month == "January"]), mean(myDF$hr170_220_minutes[myDF$month == "January"]),
mean(myDF$average_resting_hr[myDF$month == "January"]), mean(myDF$bmi[myDF$month == "January"]),
mean(myDF$minutes_asleep[myDF$month == "January"]), mean(myDF$sleep_efficiency[myDF$month == "January"]),
mean(myDF$bmi[myDF$weight == "January"])
),
Feb = c(mean(myDF$steps[myDF$month == "February"]), mean(myDF$floors_climbed[myDF$month == "February"]),
mean(myDF$total_miles[myDF$month == "February"]), mean(myDF$lightly_active_miles[myDF$month == "February"]),
mean(myDF$moderately_active_miles[myDF$month == "February"]), mean(myDF$very_active_miles[myDF$month == "February"]),
mean(myDF$sedentary_minutes[myDF$month == "February"]), mean(myDF$lightly_active_minutes[myDF$month == "February"]),
mean(myDF$fairly_active_minutes[myDF$month == "February"]), mean(myDF$very_active_minutes[myDF$month == "February"]),
mean(myDF$hr30_100_minutes[myDF$month == "February"]), mean(myDF$hr100_140_minutes[myDF$month == "February"]),
mean(myDF$hr140_170_minutes[myDF$month == "February"]), mean(myDF$hr170_220_minutes[myDF$month == "February"]),
mean(myDF$average_resting_hr[myDF$month == "February"]), mean(myDF$bmi[myDF$month == "February"]),
mean(myDF$minutes_asleep[myDF$month == "February"]), mean(myDF$sleep_efficiency[myDF$month == "February"]),
mean(myDF$bmi[myDF$weight == "February"])
),
Mar = c(mean(myDF$steps[myDF$month == "March"]), mean(myDF$floors_climbed[myDF$month == "March"]),
mean(myDF$total_miles[myDF$month == "March"]), mean(myDF$lightly_active_miles[myDF$month == "March"]),
mean(myDF$moderately_active_miles[myDF$month == "March"]), mean(myDF$very_active_miles[myDF$month == "March"]),
mean(myDF$sedentary_minutes[myDF$month == "March"]), mean(myDF$lightly_active_minutes[myDF$month == "March"]),
mean(myDF$fairly_active_minutes[myDF$month == "March"]), mean(myDF$very_active_minutes[myDF$month == "March"]),
mean(myDF$hr30_100_minutes[myDF$month == "March"]), mean(myDF$hr100_140_minutes[myDF$month == "March"]),
mean(myDF$hr140_170_minutes[myDF$month == "March"]), mean(myDF$hr170_220_minutes[myDF$month == "March"]),
mean(myDF$average_resting_hr[myDF$month == "March"]), mean(myDF$bmi[myDF$month == "March"]),
mean(myDF$minutes_asleep[myDF$month == "March"]), mean(myDF$sleep_efficiency[myDF$month == "March"]),
mean(myDF$bmi[myDF$weight == "March"])
),
Apr = c(mean(myDF$steps[myDF$month == "April"]), mean(myDF$floors_climbed[myDF$month == "April"]),
mean(myDF$total_miles[myDF$month == "April"]), mean(myDF$lightly_active_miles[myDF$month == "April"]),
mean(myDF$moderately_active_miles[myDF$month == "April"]), mean(myDF$very_active_miles[myDF$month == "April"]),
mean(myDF$sedentary_minutes[myDF$month == "April"]), mean(myDF$lightly_active_minutes[myDF$month == "April"]),
mean(myDF$fairly_active_minutes[myDF$month == "April"]), mean(myDF$very_active_minutes[myDF$month == "April"]),
mean(myDF$hr30_100_minutes[myDF$month == "April"]), mean(myDF$hr100_140_minutes[myDF$month == "April"]),
mean(myDF$hr140_170_minutes[myDF$month == "April"]), mean(myDF$hr170_220_minutes[myDF$month == "April"]),
mean(myDF$average_resting_hr[myDF$month == "April"]), mean(myDF$bmi[myDF$month == "April"]),
mean(myDF$minutes_asleep[myDF$month == "April"]), mean(myDF$sleep_efficiency[myDF$month == "April"]),
mean(myDF$bmi[myDF$weight == "April"])
),
May = c(mean(myDF$steps[myDF$month == "May"]), mean(myDF$floors_climbed[myDF$month == "May"]),
mean(myDF$total_miles[myDF$month == "May"]), mean(myDF$lightly_active_miles[myDF$month == "May"]),
mean(myDF$moderately_active_miles[myDF$month == "May"]), mean(myDF$very_active_miles[myDF$month == "May"]),
mean(myDF$sedentary_minutes[myDF$month == "May"]), mean(myDF$lightly_active_minutes[myDF$month == "May"]),
mean(myDF$fairly_active_minutes[myDF$month == "May"]), mean(myDF$very_active_minutes[myDF$month == "May"]),
mean(myDF$hr30_100_minutes[myDF$month == "May"]), mean(myDF$hr100_140_minutes[myDF$month == "May"]),
mean(myDF$hr140_170_minutes[myDF$month == "May"]), mean(myDF$hr170_220_minutes[myDF$month == "May"]),
mean(myDF$average_resting_hr[myDF$month == "May"]), mean(myDF$bmi[myDF$month == "May"]),
mean(myDF$minutes_asleep[myDF$month == "May"]), mean(myDF$sleep_efficiency[myDF$month == "May"]),
mean(myDF$bmi[myDF$weight == "May"])
),
Jun = c(mean(myDF$steps[myDF$month == "June"]), mean(myDF$floors_climbed[myDF$month == "June"]),
mean(myDF$total_miles[myDF$month == "June"]), mean(myDF$lightly_active_miles[myDF$month == "June"]),
mean(myDF$moderately_active_miles[myDF$month == "June"]), mean(myDF$very_active_miles[myDF$month == "June"]),
mean(myDF$sedentary_minutes[myDF$month == "June"]), mean(myDF$lightly_active_minutes[myDF$month == "June"]),
mean(myDF$fairly_active_minutes[myDF$month == "June"]), mean(myDF$very_active_minutes[myDF$month == "June"]),
mean(myDF$hr30_100_minutes[myDF$month == "June"]), mean(myDF$hr100_140_minutes[myDF$month == "June"]),
mean(myDF$hr140_170_minutes[myDF$month == "June"]), mean(myDF$hr170_220_minutes[myDF$month == "June"]),
mean(myDF$average_resting_hr[myDF$month == "June"]), mean(myDF$bmi[myDF$month == "June"]),
mean(myDF$minutes_asleep[myDF$month == "June"]), mean(myDF$sleep_efficiency[myDF$month == "June"]),
mean(myDF$bmi[myDF$weight == "June"])
),
Jul = c(mean(myDF$steps[myDF$month == "July"]), mean(myDF$floors_climbed[myDF$month == "July"]),
mean(myDF$total_miles[myDF$month == "July"]), mean(myDF$lightly_active_miles[myDF$month == "July"]),
mean(myDF$moderately_active_miles[myDF$month == "July"]), mean(myDF$very_active_miles[myDF$month == "July"]),
mean(myDF$sedentary_minutes[myDF$month == "July"]), mean(myDF$lightly_active_minutes[myDF$month == "July"]),
mean(myDF$fairly_active_minutes[myDF$month == "July"]), mean(myDF$very_active_minutes[myDF$month == "July"]),
mean(myDF$hr30_100_minutes[myDF$month == "July"]), mean(myDF$hr100_140_minutes[myDF$month == "July"]),
mean(myDF$hr140_170_minutes[myDF$month == "July"]), mean(myDF$hr170_220_minutes[myDF$month == "July"]),
mean(myDF$average_resting_hr[myDF$month == "July"]), mean(myDF$bmi[myDF$month == "July"]),
mean(myDF$minutes_asleep[myDF$month == "July"]), mean(myDF$sleep_efficiency[myDF$month == "July"]),
mean(myDF$bmi[myDF$weight == "July"])
),
Aug = c(mean(myDF$steps[myDF$month == "August"]), mean(myDF$floors_climbed[myDF$month == "August"]),
mean(myDF$total_miles[myDF$month == "August"]), mean(myDF$lightly_active_miles[myDF$month == "August"]),
mean(myDF$moderately_active_miles[myDF$month == "August"]), mean(myDF$very_active_miles[myDF$month == "August"]),
mean(myDF$sedentary_minutes[myDF$month == "August"]), mean(myDF$lightly_active_minutes[myDF$month == "August"]),
mean(myDF$fairly_active_minutes[myDF$month == "August"]), mean(myDF$very_active_minutes[myDF$month == "August"]),
mean(myDF$hr30_100_minutes[myDF$month == "August"]), mean(myDF$hr100_140_minutes[myDF$month == "August"]),
mean(myDF$hr140_170_minutes[myDF$month == "August"]), mean(myDF$hr170_220_minutes[myDF$month == "August"]),
mean(myDF$average_resting_hr[myDF$month == "August"]), mean(myDF$bmi[myDF$month == "August"]),
mean(myDF$minutes_asleep[myDF$month == "August"]), mean(myDF$sleep_efficiency[myDF$month == "August"]),
mean(myDF$bmi[myDF$weight == "August"])
),
Sep = c(mean(myDF$steps[myDF$month == "September"]), mean(myDF$floors_climbed[myDF$month == "September"]),
mean(myDF$total_miles[myDF$month == "September"]), mean(myDF$lightly_active_miles[myDF$month == "September"]),
mean(myDF$moderately_active_miles[myDF$month == "September"]), mean(myDF$very_active_miles[myDF$month == "September"]),
mean(myDF$sedentary_minutes[myDF$month == "September"]), mean(myDF$lightly_active_minutes[myDF$month == "September"]),
mean(myDF$fairly_active_minutes[myDF$month == "September"]), mean(myDF$very_active_minutes[myDF$month == "September"]),
mean(myDF$hr30_100_minutes[myDF$month == "September"]), mean(myDF$hr100_140_minutes[myDF$month == "September"]),
mean(myDF$hr140_170_minutes[myDF$month == "September"]), mean(myDF$hr170_220_minutes[myDF$month == "September"]),
mean(myDF$average_resting_hr[myDF$month == "September"]), mean(myDF$bmi[myDF$month == "September"]),
mean(myDF$minutes_asleep[myDF$month == "September"]), mean(myDF$sleep_efficiency[myDF$month == "September"]),
mean(myDF$bmi[myDF$weight == "September"])
),
Oct = c(mean(myDF$steps[myDF$month == "October"]), mean(myDF$floors_climbed[myDF$month == "October"]),
mean(myDF$total_miles[myDF$month == "October"]), mean(myDF$lightly_active_miles[myDF$month == "October"]),
mean(myDF$moderately_active_miles[myDF$month == "October"]), mean(myDF$very_active_miles[myDF$month == "October"]),
mean(myDF$sedentary_minutes[myDF$month == "October"]), mean(myDF$lightly_active_minutes[myDF$month == "October"]),
mean(myDF$fairly_active_minutes[myDF$month == "October"]), mean(myDF$very_active_minutes[myDF$month == "October"]),
mean(myDF$hr30_100_minutes[myDF$month == "October"]), mean(myDF$hr100_140_minutes[myDF$month == "October"]),
mean(myDF$hr140_170_minutes[myDF$month == "October"]), mean(myDF$hr170_220_minutes[myDF$month == "October"]),
mean(myDF$average_resting_hr[myDF$month == "October"]), mean(myDF$bmi[myDF$month == "October"]),
mean(myDF$minutes_asleep[myDF$month == "October"]), mean(myDF$sleep_efficiency[myDF$month == "October"]),
mean(myDF$bmi[myDF$weight == "October"])
),
Nov = c(mean(myDF$steps[myDF$month == "November"]), mean(myDF$floors_climbed[myDF$month == "November"]),
mean(myDF$total_miles[myDF$month == "November"]), mean(myDF$lightly_active_miles[myDF$month == "November"]),
mean(myDF$moderately_active_miles[myDF$month == "November"]), mean(myDF$very_active_miles[myDF$month == "November"]),
mean(myDF$sedentary_minutes[myDF$month == "November"]), mean(myDF$lightly_active_minutes[myDF$month == "November"]),
mean(myDF$fairly_active_minutes[myDF$month == "November"]), mean(myDF$very_active_minutes[myDF$month == "November"]),
mean(myDF$hr30_100_minutes[myDF$month == "November"]), mean(myDF$hr100_140_minutes[myDF$month == "November"]),
mean(myDF$hr140_170_minutes[myDF$month == "November"]), mean(myDF$hr170_220_minutes[myDF$month == "November"]),
mean(myDF$average_resting_hr[myDF$month == "November"]), mean(myDF$bmi[myDF$month == "November"]),
mean(myDF$minutes_asleep[myDF$month == "November"]), mean(myDF$sleep_efficiency[myDF$month == "November"]),
mean(myDF$bmi[myDF$weight == "November"])
),
Dec = c(mean(myDF$steps[myDF$month == "December"]), mean(myDF$floors_climbed[myDF$month == "December"]),
mean(myDF$total_miles[myDF$month == "December"]), mean(myDF$lightly_active_miles[myDF$month == "December"]),
mean(myDF$moderately_active_miles[myDF$month == "December"]), mean(myDF$very_active_miles[myDF$month == "December"]),
mean(myDF$sedentary_minutes[myDF$month == "December"]), mean(myDF$lightly_active_minutes[myDF$month == "December"]),
mean(myDF$fairly_active_minutes[myDF$month == "December"]), mean(myDF$very_active_minutes[myDF$month == "December"]),
mean(myDF$hr30_100_minutes[myDF$month == "December"]), mean(myDF$hr100_140_minutes[myDF$month == "December"]),
mean(myDF$hr140_170_minutes[myDF$month == "December"]), mean(myDF$hr170_220_minutes[myDF$month == "December"]),
mean(myDF$average_resting_hr[myDF$month == "December"]), mean(myDF$bmi[myDF$month == "December"]),
mean(myDF$minutes_asleep[myDF$month == "December"]), mean(myDF$sleep_efficiency[myDF$month == "December"]),
mean(myDF$bmi[myDF$weight == "December"])
)
)
#create the averages data frame that contains average statistics per weekday
averagesDF2 <- data.frame(
Var = c("Steps", "Floors Climbed", "Total Miles", "Lightly Active Miles", "Moderately Active Miles",
"Very Active Miles", "Sedentary Minutes", "Lightly Active Minutes", "Fairly Active Minutes",
"Very Active Minutes", "HR 30-100 Minutes", "HR 100-140 Minutes", "HR 140-170 Minutes",
"HR 170-220 Minutes", "Average Resting HR", "BMI", "Minutes Asleep", "Sleep Efficiency", "Weight"),
Mon = c(mean(myDF$steps[myDF$day_of_week == "Monday"]), mean(myDF$floors_climbed[myDF$day_of_week == "Monday"]),
mean(myDF$total_miles[myDF$day_of_week == "Monday"]), mean(myDF$lightly_active_miles[myDF$day_of_week == "Monday"]),
mean(myDF$moderately_active_miles[myDF$day_of_week == "Monday"]), mean(myDF$very_active_miles[myDF$day_of_week == "Monday"]),
mean(myDF$sedentary_minutes[myDF$day_of_week == "Monday"]), mean(myDF$lightly_active_minutes[myDF$day_of_week == "Monday"]),
mean(myDF$fairly_active_minutes[myDF$day_of_week == "Monday"]), mean(myDF$very_active_minutes[myDF$day_of_week == "Monday"]),
mean(myDF$hr30_100_minutes[myDF$month == "Monday"]), mean(myDF$hr100_140_minutes[myDF$month == "Monday"]),
mean(myDF$hr140_170_minutes[myDF$month == "Monday"]), mean(myDF$hr170_220_minutes[myDF$month == "Monday"]),
mean(myDF$average_resting_hr[myDF$month == "Monday"]), mean(myDF$bmi[myDF$month == "Monday"]),
mean(myDF$minutes_asleep[myDF$month == "Monday"]), mean(myDF$sleep_efficiency[myDF$month == "Monday"]),
mean(myDF$bmi[myDF$weight == "Monday"])
),
Tue = c(mean(myDF$steps[myDF$day_of_week == "Tuesday"]), mean(myDF$floors_climbed[myDF$day_of_week == "Tuesday"]),
mean(myDF$total_miles[myDF$day_of_week == "Tuesday"]), mean(myDF$lightly_active_miles[myDF$day_of_week == "Tuesday"]),
mean(myDF$moderately_active_miles[myDF$day_of_week == "Tuesday"]), mean(myDF$very_active_miles[myDF$day_of_week == "Tuesday"]),
mean(myDF$sedentary_minutes[myDF$day_of_week == "Tuesday"]), mean(myDF$lightly_active_minutes[myDF$day_of_week == "Tuesday"]),
mean(myDF$fairly_active_minutes[myDF$day_of_week == "Tuesday"]), mean(myDF$very_active_minutes[myDF$day_of_week == "Tuesday"]),
mean(myDF$hr30_100_minutes[myDF$month == "Tuesday"]), mean(myDF$hr100_140_minutes[myDF$month == "Tuesday"]),
mean(myDF$hr140_170_minutes[myDF$month == "Tuesday"]), mean(myDF$hr170_220_minutes[myDF$month == "Tuesday"]),
mean(myDF$average_resting_hr[myDF$month == "Tuesday"]), mean(myDF$bmi[myDF$month == "Tuesday"]),
mean(myDF$minutes_asleep[myDF$month == "Tuesday"]), mean(myDF$sleep_efficiency[myDF$month == "Tuesday"]),
mean(myDF$bmi[myDF$weight == "Tuesday"])
),
Wed = c(mean(myDF$steps[myDF$day_of_week == "Wednesday"]), mean(myDF$floors_climbed[myDF$day_of_week == "Wednesday"]),
mean(myDF$total_miles[myDF$day_of_week == "Wednesday"]), mean(myDF$lightly_active_miles[myDF$day_of_week == "Wednesday"]),
mean(myDF$moderately_active_miles[myDF$day_of_week == "Wednesday"]), mean(myDF$very_active_miles[myDF$day_of_week == "Wednesday"]),
mean(myDF$sedentary_minutes[myDF$day_of_week == "Wednesday"]), mean(myDF$lightly_active_minutes[myDF$day_of_week == "Wednesday"]),
mean(myDF$fairly_active_minutes[myDF$day_of_week == "Wednesday"]), mean(myDF$very_active_minutes[myDF$day_of_week == "Wednesday"]),
mean(myDF$hr30_100_minutes[myDF$month == "Wednesday"]), mean(myDF$hr100_140_minutes[myDF$month == "Wednesday"]),
mean(myDF$hr140_170_minutes[myDF$month == "Wednesday"]), mean(myDF$hr170_220_minutes[myDF$month == "Wednesday"]),
mean(myDF$average_resting_hr[myDF$month == "Wednesday"]), mean(myDF$bmi[myDF$month == "Wednesday"]),
mean(myDF$minutes_asleep[myDF$month == "Wednesday"]), mean(myDF$sleep_efficiency[myDF$month == "Wednesday"]),
mean(myDF$bmi[myDF$weight == "Wednesday"])
),
Thu = c(mean(myDF$steps[myDF$day_of_week == "Thursday"]), mean(myDF$floors_climbed[myDF$day_of_week == "Thursday"]),
mean(myDF$total_miles[myDF$day_of_week == "Thursday"]), mean(myDF$lightly_active_miles[myDF$day_of_week == "Thursday"]),
mean(myDF$moderately_active_miles[myDF$day_of_week == "Thursday"]), mean(myDF$very_active_miles[myDF$day_of_week == "Thursday"]),
mean(myDF$sedentary_minutes[myDF$day_of_week == "Thursday"]), mean(myDF$lightly_active_minutes[myDF$day_of_week == "Thursday"]),
mean(myDF$fairly_active_minutes[myDF$day_of_week == "Thursday"]), mean(myDF$very_active_minutes[myDF$day_of_week == "Thursday"]),
mean(myDF$hr30_100_minutes[myDF$month == "Thursday"]), mean(myDF$hr100_140_minutes[myDF$month == "Thursday"]),
mean(myDF$hr140_170_minutes[myDF$month == "Thursday"]), mean(myDF$hr170_220_minutes[myDF$month == "Thursday"]),
mean(myDF$average_resting_hr[myDF$month == "Thursday"]), mean(myDF$bmi[myDF$month == "Thursday"]),
mean(myDF$minutes_asleep[myDF$month == "Thursday"]), mean(myDF$sleep_efficiency[myDF$month == "Thursday"]),
mean(myDF$bmi[myDF$weight == "Thursday"])
),
Fri = c(mean(myDF$steps[myDF$day_of_week == "Friday"]), mean(myDF$floors_climbed[myDF$day_of_week == "Friday"]),
mean(myDF$total_miles[myDF$day_of_week == "Friday"]), mean(myDF$lightly_active_miles[myDF$day_of_week == "Friday"]),
mean(myDF$moderately_active_miles[myDF$day_of_week == "Friday"]), mean(myDF$very_active_miles[myDF$day_of_week == "Friday"]),
mean(myDF$sedentary_minutes[myDF$day_of_week == "Friday"]), mean(myDF$lightly_active_minutes[myDF$day_of_week == "Friday"]),
mean(myDF$fairly_active_minutes[myDF$day_of_week == "Friday"]), mean(myDF$very_active_minutes[myDF$day_of_week == "Friday"]),
mean(myDF$hr30_100_minutes[myDF$month == "Friday"]), mean(myDF$hr100_140_minutes[myDF$month == "Friday"]),
mean(myDF$hr140_170_minutes[myDF$month == "Friday"]), mean(myDF$hr170_220_minutes[myDF$month == "Friday"]),
mean(myDF$average_resting_hr[myDF$month == "Friday"]), mean(myDF$bmi[myDF$month == "Friday"]),
mean(myDF$minutes_asleep[myDF$month == "Friday"]), mean(myDF$sleep_efficiency[myDF$month == "Friday"]),
mean(myDF$bmi[myDF$weight == "Friday"])
),
Sat = c(mean(myDF$steps[myDF$day_of_week == "Saturday"]), mean(myDF$floors_climbed[myDF$day_of_week == "Saturday"]),
mean(myDF$total_miles[myDF$day_of_week == "Saturday"]), mean(myDF$lightly_active_miles[myDF$day_of_week == "Saturday"]),
mean(myDF$moderately_active_miles[myDF$day_of_week == "Saturday"]), mean(myDF$very_active_miles[myDF$day_of_week == "Saturday"]),
mean(myDF$sedentary_minutes[myDF$day_of_week == "Saturday"]), mean(myDF$lightly_active_minutes[myDF$day_of_week == "Saturday"]),
mean(myDF$fairly_active_minutes[myDF$day_of_week == "Saturday"]), mean(myDF$very_active_minutes[myDF$day_of_week == "Saturday"]),
mean(myDF$hr30_100_minutes[myDF$month == "Saturday"]), mean(myDF$hr100_140_minutes[myDF$month == "Saturday"]),
mean(myDF$hr140_170_minutes[myDF$month == "Saturday"]), mean(myDF$hr170_220_minutes[myDF$month == "Saturday"]),
mean(myDF$average_resting_hr[myDF$month == "Saturday"]), mean(myDF$bmi[myDF$month == "Saturday"]),
mean(myDF$minutes_asleep[myDF$month == "Saturday"]), mean(myDF$sleep_efficiency[myDF$month == "Saturday"]),
mean(myDF$bmi[myDF$weight == "Saturday"])
),
Sun = c(mean(myDF$steps[myDF$day_of_week == "Sunday"]), mean(myDF$floors_climbed[myDF$day_of_week == "Sunday"]),
mean(myDF$total_miles[myDF$day_of_week == "Sunday"]), mean(myDF$lightly_active_miles[myDF$day_of_week == "Sunday"]),
mean(myDF$moderately_active_miles[myDF$day_of_week == "Sunday"]), mean(myDF$very_active_miles[myDF$day_of_week == "Sunday"]),
mean(myDF$sedentary_minutes[myDF$day_of_week == "Sunday"]), mean(myDF$lightly_active_minutes[myDF$day_of_week == "Sunday"]),
mean(myDF$fairly_active_minutes[myDF$day_of_week == "Sunday"]), mean(myDF$very_active_minutes[myDF$day_of_week == "Sunday"]),
mean(myDF$hr30_100_minutes[myDF$month == "Sunday"]), mean(myDF$hr100_140_minutes[myDF$month == "Sunday"]),
mean(myDF$hr140_170_minutes[myDF$month == "Sunday"]), mean(myDF$hr170_220_minutes[myDF$month == "Sunday"]),
mean(myDF$average_resting_hr[myDF$month == "Sunday"]), mean(myDF$bmi[myDF$month == "Sunday"]),
mean(myDF$minutes_asleep[myDF$month == "Sunday"]), mean(myDF$sleep_efficiency[myDF$month == "Sunday"]),
mean(myDF$bmi[myDF$weight == "Sunday"])
)
)
#theme of dashboard set to flatly
ui<-fluidPage(theme = shinytheme("flatly"),
navbarPage("Fitbit Dashboard",
# tabPanel 1
# first panel
tabPanel("How to Use",
img(src = "https://assets.phenompeople.com/CareerConnectResources/MERCUS/social/1200x630-1552902977090.jpg", height = 120, width = 220),
h1("Biometric Data Dashboard"),
mainPanel(
strong("Refer to the website here for more information:"),
a("https://www.merck.com/index.html"),
h3("How to Use My Dashboard"),
htmlOutput("text2")
)
),
# tabPanel 2
tabPanel("Activity Progress",
sidebarLayout(
sidebarPanel(h3("Activity Progress"),
dateRangeInput("dates", ("Date range")),
varSelectInput("variables1", ("Variable:"), myDF[c(-1, -2)]),
p("Please select N/A on the stacked variables scrolldown to",
"view single variable data"), br(),
selectInput("stack", ("Variable (Stacked):"),
c("N/A" = "nana", "Active Miles" = "miles", "Active minutes" = "minutes", "HR minutes" = "hr")),
),
mainPanel(
em("Select a date range and a variable!"),
strong("Refer to the website here for more information:"),
a("https://www.merck.com/index.html"),
plotOutput(outputId = "bioBarGraph")
)
)
),
# tab Panel 3
tabPanel("Average Statistics",
sidebarLayout(
sidebarPanel(br(), br(), br(),
h3("Average Daily Statistics Per Month"),
p("Pick a variable to view the average daily amount of steps,
floors climbed... etc taken each month."),
br(), br(),
selectizeInput("variables", "Variables (Month)", choices = unique(averagesDF$Var), selected = unique(averagesDF$Var)[1]),
br(), br(), br(), br(), br(), br(), br(), br(), br(),
h3("Average Daily Statistics Per Weekday"),
p("Pick a variable to view the average daily amount of steps,
floors climbed... etc taken each week."),
br(), br(),
selectizeInput("variables2", "Variables (Weekdays)", choices = unique(averagesDF2$Var), selected = unique(averagesDF2$Var)[1]),
br(), br(), br(), br()
),
mainPanel(
highchartOutput("plotAvgs"),
highchartOutput("plotAvgs2")
)
)
),
#tabPanel 4
tabPanel("Data Table",
sidebarLayout(
sidebarPanel(
conditionalPanel(
'input.dataset === "myDF"',
checkboxGroupInput("show_vars", "Columns of information to show:",
names(myDF), selected = list("collection_date", "steps"))
),
),
mainPanel(
tabsetPanel(
id = 'dataset',
tabPanel("myDF", DT::dataTableOutput("mytable1"))
)
)
)
)
)
)
server <- function(input, output) {
# You can access the values of the widget (as a vector of Dates)
# with input$collectiondates,
# e.g. output$value <- renderPrint({ input$dates })
# text on tabPanel 1
output$text2 <- renderUI({
HTML(paste("", "1. Activity Progress",
"- Date Range: select a start and an end date to set a date range of the activity progress",
"- Variables: use the drop-down to select a variable of the activity progress you wish to see",
"- Variables (Stacked): use the drop-down to select a variable of the activity progress you wish to see as a stacked bar chart",
"", "2. Average Statistics",
"- Displays the average daily statistics per month or day of week",
"- Monthly Data: select which variable you want to see average statistics for",
"- Weekday Data: select which variable you want to see average statistics for",
"", "3. Data Table ",
"- Select which variables you want visible for each data point",
"- Able to filter through data using filter as well as placing them in order by any variable",
"- Steps and Data are pre chosen variables",
sep="<br/>"))
})
# choose columns to display
myDF2 = myDF[sample(nrow(myDF), nrow(myDF), replace = FALSE), ]
output$mytable1 <- DT::renderDataTable({
DT::datatable(myDF2[, input$show_vars, drop = FALSE], rownames = FALSE, filter = 'top')
})
# sorted columns are colored now because CSS are attached to them
#output$mytable2 <- DT::renderDataTable({
# DT::datatable(finalDataFrame, options = list(orderClasses = TRUE), rownames = FALSE)
#})
#creates bar plot using monthly averages data frame
reactivedf <- reactive({
filtereddf <- averagesDF %>%
#chooses what variable needs to be displayed
dplyr::filter(Var == input$variables)
filtereddf
})
#creates bar plot using weekly averages data frame
reactivedf2 <- reactive({
filtereddf2 <- averagesDF2 %>%
#chooses what variable needs to be displayed
dplyr::filter(Var == input$variables2)
filtereddf2
})
output$bioBarGraph <- renderPlot({
# counter for the start date - date range
counter1 = 0
for (myDate in as.character(myDF$collection_date)){
counter1 = counter1 + 1
if (myDate == input$dates[1]) {
break
}
}
# counter for the end date - date range
counter2 = 0
for (myDate in as.character(myDF$collection_date)){
counter2 = counter2 + 1
if (myDate == input$dates[2]) {
break
}
}
# subset of myDF with the selected date range
mySubset <- myDF[c(counter1:counter2),]
if (input$stack == "nana"){ # single bar chart - if stacked bar is N/A
ggplot(data=mySubset, aes(x=collection_date, y=!!input$variables1)) + geom_bar(stat="identity", width = .5, fill = "#13A999")
}else if(input$stack == "miles"){ # stacked bar chart - active miles
ggplot(data=miles, aes(x=collection_date, y=value, fill=variable)) + geom_bar(stat="identity", width = .5)
}else if(input$stack == "minutes"){ # stacked bar chart - active minutes
ggplot(data=minutes, aes(x=collection_date, y=value, fill=variable)) + geom_bar(stat="identity", width = .5)
}else if(input$stack == "hr"){ # stacked bar chart - HR minutes
ggplot(data=hr, aes(x=collection_date, y=value, fill=variable)) + geom_bar(stat="identity", width = .5)
}
})
#actually outputs the plot object
output$plotAvgs <- renderHighchart({
highchart() %>%
#each column made for each month of the year
hc_add_series(type = "column", reactivedf()$Jan, name = "Jan") %>%
hc_add_series(type = "column", reactivedf()$Feb, name = "Feb") %>%
hc_add_series(type = "column", reactivedf()$Mar, name = "Mar") %>%
hc_add_series(type = "column", reactivedf()$Apr, name = "Apr") %>%
hc_add_series(type = "column", reactivedf()$May, name = "May") %>%
hc_add_series(type = "column", reactivedf()$Jun, name = "Jun") %>%
hc_add_series(type = "column", reactivedf()$Jul, name = "Jul") %>%
hc_add_series(type = "column", reactivedf()$Aug, name = "Aug") %>%
hc_add_series(type = "column", reactivedf()$Sep, name = "Sep") %>%
hc_add_series(type = "column", reactivedf()$Oct, name = "Oct") %>%
hc_add_series(type = "column", reactivedf()$Nov, name = "Nov") %>%
hc_add_series(type = "column", reactivedf()$Dec, name = "Dec") %>%
hc_xAxis(labels = list(enabled = FALSE)) %>%
hc_title(text = input$variables)
})
#actually outputs the plot object
output$plotAvgs2 <- renderHighchart({
highchart() %>%
#each column made for each day of the week
hc_add_series(type = "column", reactivedf2()$Mon, name = "Mon") %>%
hc_add_series(type = "column", reactivedf2()$Tue, name = "Tue") %>%
hc_add_series(type = "column", reactivedf2()$Wed, name = "Wed") %>%
hc_add_series(type = "column", reactivedf2()$Thu, name = "Thu") %>%
hc_add_series(type = "column", reactivedf2()$Fri, name = "Fri") %>%
hc_add_series(type = "column", reactivedf2()$Sat, name = "Sat") %>%
hc_add_series(type = "column", reactivedf2()$Sun, name = "Sun") %>%
hc_xAxis(labels = list(enabled = FALSE)) %>%
hc_title(text = input$variables2)
})
}
shinyApp(ui, server)8.16.2 Total Patients Visualization Code
#ALL PATIENTS
library(RMariaDB)
library(shiny)
library(ggplot2)
library(reshape2)
library(highcharter)
library(dplyr)
library(shinythemes)
library(DT)
# connect to the database
db <- dbConnect(RMariaDB::MariaDB(),
host = "scholar-db.rcac.purdue.edu",
db = "merck",
user = "merck_user",
password = "Ph$rma_user")
#get the correct data from the database
myDF <- dbGetQuery(db, "SELECT * FROM fitbit_data;")
#changes format of the collection date so that it can be formatted
myDF$collection_date <- sapply(myDF$collection_date, substring, 0, 10)
# active miles (stacked)
miles1 <- subset(myDF, select = c('collection_date','lightly_active_miles', 'moderately_active_miles', 'very_active_miles'))
miles <- melt(miles1)
miles$collection_date <- as.Date(miles$collection_date)
# active minutes (stacked)
minutes1 <- subset(myDF, select = c('collection_date', 'sedentary_minutes', 'lightly_active_minutes', 'fairly_active_minutes', 'very_active_minutes'))
minutes <- melt(minutes1)
minutes$collection_date <- as.Date(minutes$collection_date)
# HR minutes (stacked)
hr1 <- subset(myDF, select = c('collection_date', 'hr30_100_minutes', 'hr100_140_minutes', 'hr140_170_minutes', 'hr170_220_minutes'))
hr <- melt(hr1)
hr$collection_date <- as.Date(hr$collection_date)
# myDF to date format
myDF$collection_date = as.Date(myDF$collection_date)
myDF$year <- format(myDF$collection_date,'%Y')
myDF$month <- factor(format(myDF$collection_date,'%B'), levels = month.name)
myDF$day_of_week <- weekdays(myDF$collection_date)
#creates a data frame that finds average statistics of each month
averagesDF <- data.frame(
Var = c("Steps", "Total Miles", "Lightly Active Miles", "Moderately Active Miles",
"Very Active Miles", "Sedentary Minutes", "Lightly Active Minutes", "Fairly Active Minutes",
"Very Active Minutes"),
Jan = c(mean(myDF$steps[myDF$month == "January"]),
mean(myDF$total_miles[myDF$month == "January"]), mean(myDF$lightly_active_miles[myDF$month == "January"]),
mean(myDF$moderately_active_miles[myDF$month == "January"]), mean(myDF$very_active_miles[myDF$month == "January"]),
mean(myDF$sedentary_minutes[myDF$month == "January"]), mean(myDF$lightly_active_minutes[myDF$month == "January"]),
mean(myDF$fairly_active_minutes[myDF$month == "January"]), mean(myDF$very_active_minutes[myDF$month == "January"])
),
Feb = c(mean(myDF$steps[myDF$month == "February"]),
mean(myDF$total_miles[myDF$month == "February"]), mean(myDF$lightly_active_miles[myDF$month == "February"]),
mean(myDF$moderately_active_miles[myDF$month == "February"]), mean(myDF$very_active_miles[myDF$month == "February"]),
mean(myDF$sedentary_minutes[myDF$month == "February"]), mean(myDF$lightly_active_minutes[myDF$month == "February"]),
mean(myDF$fairly_active_minutes[myDF$month == "February"]), mean(myDF$very_active_minutes[myDF$month == "February"])
),
Mar = c(mean(myDF$steps[myDF$month == "March"]),
mean(myDF$total_miles[myDF$month == "March"]), mean(myDF$lightly_active_miles[myDF$month == "March"]),
mean(myDF$moderately_active_miles[myDF$month == "March"]), mean(myDF$very_active_miles[myDF$month == "March"]),
mean(myDF$sedentary_minutes[myDF$month == "March"]), mean(myDF$lightly_active_minutes[myDF$month == "March"]),
mean(myDF$fairly_active_minutes[myDF$month == "March"]), mean(myDF$very_active_minutes[myDF$month == "March"])
),
Apr = c(mean(myDF$steps[myDF$month == "April"]),
mean(myDF$total_miles[myDF$month == "April"]), mean(myDF$lightly_active_miles[myDF$month == "April"]),
mean(myDF$moderately_active_miles[myDF$month == "April"]), mean(myDF$very_active_miles[myDF$month == "April"]),
mean(myDF$sedentary_minutes[myDF$month == "April"]), mean(myDF$lightly_active_minutes[myDF$month == "April"]),
mean(myDF$fairly_active_minutes[myDF$month == "April"]), mean(myDF$very_active_minutes[myDF$month == "April"])
),
May = c(mean(myDF$steps[myDF$month == "May"]),
mean(myDF$total_miles[myDF$month == "May"]), mean(myDF$lightly_active_miles[myDF$month == "May"]),
mean(myDF$moderately_active_miles[myDF$month == "May"]), mean(myDF$very_active_miles[myDF$month == "May"]),
mean(myDF$sedentary_minutes[myDF$month == "May"]), mean(myDF$lightly_active_minutes[myDF$month == "May"]),
mean(myDF$fairly_active_minutes[myDF$month == "May"]), mean(myDF$very_active_minutes[myDF$month == "May"])
),
Jun = c(mean(myDF$steps[myDF$month == "June"]),
mean(myDF$total_miles[myDF$month == "June"]), mean(myDF$lightly_active_miles[myDF$month == "June"]),
mean(myDF$moderately_active_miles[myDF$month == "June"]), mean(myDF$very_active_miles[myDF$month == "June"]),
mean(myDF$sedentary_minutes[myDF$month == "June"]), mean(myDF$lightly_active_minutes[myDF$month == "June"]),
mean(myDF$fairly_active_minutes[myDF$month == "June"]), mean(myDF$very_active_minutes[myDF$month == "June"])
),
Jul = c(mean(myDF$steps[myDF$month == "July"]),
mean(myDF$total_miles[myDF$month == "July"]), mean(myDF$lightly_active_miles[myDF$month == "July"]),
mean(myDF$moderately_active_miles[myDF$month == "July"]), mean(myDF$very_active_miles[myDF$month == "July"]),
mean(myDF$sedentary_minutes[myDF$month == "July"]), mean(myDF$lightly_active_minutes[myDF$month == "July"]),
mean(myDF$fairly_active_minutes[myDF$month == "July"]), mean(myDF$very_active_minutes[myDF$month == "July"])
),
Aug = c(mean(myDF$steps[myDF$month == "August"]),
mean(myDF$total_miles[myDF$month == "August"]), mean(myDF$lightly_active_miles[myDF$month == "August"]),
mean(myDF$moderately_active_miles[myDF$month == "August"]), mean(myDF$very_active_miles[myDF$month == "August"]),
mean(myDF$sedentary_minutes[myDF$month == "August"]), mean(myDF$lightly_active_minutes[myDF$month == "August"]),
mean(myDF$fairly_active_minutes[myDF$month == "August"]), mean(myDF$very_active_minutes[myDF$month == "August"])
),
Sep = c(mean(myDF$steps[myDF$month == "September"]),
mean(myDF$total_miles[myDF$month == "September"]), mean(myDF$lightly_active_miles[myDF$month == "September"]),
mean(myDF$moderately_active_miles[myDF$month == "September"]), mean(myDF$very_active_miles[myDF$month == "September"]),
mean(myDF$sedentary_minutes[myDF$month == "September"]), mean(myDF$lightly_active_minutes[myDF$month == "September"]),
mean(myDF$fairly_active_minutes[myDF$month == "September"]), mean(myDF$very_active_minutes[myDF$month == "September"])
),
Oct = c(mean(myDF$steps[myDF$month == "October"]),
mean(myDF$total_miles[myDF$month == "October"]), mean(myDF$lightly_active_miles[myDF$month == "October"]),
mean(myDF$moderately_active_miles[myDF$month == "October"]), mean(myDF$very_active_miles[myDF$month == "October"]),
mean(myDF$sedentary_minutes[myDF$month == "October"]), mean(myDF$lightly_active_minutes[myDF$month == "October"]),
mean(myDF$fairly_active_minutes[myDF$month == "October"]), mean(myDF$very_active_minutes[myDF$month == "October"])
),
Nov = c(mean(myDF$steps[myDF$month == "November"]),
mean(myDF$total_miles[myDF$month == "November"]), mean(myDF$lightly_active_miles[myDF$month == "November"]),
mean(myDF$moderately_active_miles[myDF$month == "November"]), mean(myDF$very_active_miles[myDF$month == "November"]),
mean(myDF$sedentary_minutes[myDF$month == "November"]), mean(myDF$lightly_active_minutes[myDF$month == "November"]),
mean(myDF$fairly_active_minutes[myDF$month == "November"]), mean(myDF$very_active_minutes[myDF$month == "November"])
),
Dec = c(mean(myDF$steps[myDF$month == "December"]),
mean(myDF$total_miles[myDF$month == "December"]), mean(myDF$lightly_active_miles[myDF$month == "December"]),
mean(myDF$moderately_active_miles[myDF$month == "December"]), mean(myDF$very_active_miles[myDF$month == "December"]),
mean(myDF$sedentary_minutes[myDF$month == "December"]), mean(myDF$lightly_active_minutes[myDF$month == "December"]),
mean(myDF$fairly_active_minutes[myDF$month == "December"]), mean(myDF$very_active_minutes[myDF$month == "December"])
)
)
#creates a data frame that finds average statistics of each day of the week
averagesDF2 <- data.frame(
Var = c("Steps", "Total Miles", "Lightly Active Miles", "Moderately Active Miles",
"Very Active Miles", "Sedentary Minutes", "Lightly Active Minutes", "Fairly Active Minutes",
"Very Active Minutes"),
Mon = c(mean(myDF$steps[myDF$day_of_week == "Monday"]),
mean(myDF$total_miles[myDF$day_of_week == "Monday"]), mean(myDF$lightly_active_miles[myDF$day_of_week == "Monday"]),
mean(myDF$moderately_active_miles[myDF$day_of_week == "Monday"]), mean(myDF$very_active_miles[myDF$day_of_week == "Monday"]),
mean(myDF$sedentary_minutes[myDF$day_of_week == "Monday"]), mean(myDF$lightly_active_minutes[myDF$day_of_week == "Monday"]),
mean(myDF$fairly_active_minutes[myDF$day_of_week == "Monday"]), mean(myDF$very_active_minutes[myDF$day_of_week == "Monday"])
),
Tue = c(mean(myDF$steps[myDF$day_of_week == "Tuesday"]),
mean(myDF$total_miles[myDF$day_of_week == "Tuesday"]), mean(myDF$lightly_active_miles[myDF$day_of_week == "Tuesday"]),
mean(myDF$moderately_active_miles[myDF$day_of_week == "Tuesday"]), mean(myDF$very_active_miles[myDF$day_of_week == "Tuesday"]),
mean(myDF$sedentary_minutes[myDF$day_of_week == "Tuesday"]), mean(myDF$lightly_active_minutes[myDF$day_of_week == "Tuesday"]),
mean(myDF$fairly_active_minutes[myDF$day_of_week == "Tuesday"]), mean(myDF$very_active_minutes[myDF$day_of_week == "Tuesday"])
),
Wed = c(mean(myDF$steps[myDF$day_of_week == "Wednesday"]),
mean(myDF$total_miles[myDF$day_of_week == "Wednesday"]), mean(myDF$lightly_active_miles[myDF$day_of_week == "Wednesday"]),
mean(myDF$moderately_active_miles[myDF$day_of_week == "Wednesday"]), mean(myDF$very_active_miles[myDF$day_of_week == "Wednesday"]),
mean(myDF$sedentary_minutes[myDF$day_of_week == "Wednesday"]), mean(myDF$lightly_active_minutes[myDF$day_of_week == "Wednesday"]),
mean(myDF$fairly_active_minutes[myDF$day_of_week == "Wednesday"]), mean(myDF$very_active_minutes[myDF$day_of_week == "Wednesday"])
),
Thu = c(mean(myDF$steps[myDF$day_of_week == "Thursday"]),
mean(myDF$total_miles[myDF$day_of_week == "Thursday"]), mean(myDF$lightly_active_miles[myDF$day_of_week == "Thursday"]),
mean(myDF$moderately_active_miles[myDF$day_of_week == "Thursday"]), mean(myDF$very_active_miles[myDF$day_of_week == "Thursday"]),
mean(myDF$sedentary_minutes[myDF$day_of_week == "Thursday"]), mean(myDF$lightly_active_minutes[myDF$day_of_week == "Thursday"]),
mean(myDF$fairly_active_minutes[myDF$day_of_week == "Thursday"]), mean(myDF$very_active_minutes[myDF$day_of_week == "Thursday"])
),
Fri = c(mean(myDF$steps[myDF$day_of_week == "Friday"]),
mean(myDF$total_miles[myDF$day_of_week == "Friday"]), mean(myDF$lightly_active_miles[myDF$day_of_week == "Friday"]),
mean(myDF$moderately_active_miles[myDF$day_of_week == "Friday"]), mean(myDF$very_active_miles[myDF$day_of_week == "Friday"]),
mean(myDF$sedentary_minutes[myDF$day_of_week == "Friday"]), mean(myDF$lightly_active_minutes[myDF$day_of_week == "Friday"]),
mean(myDF$fairly_active_minutes[myDF$day_of_week == "Friday"]), mean(myDF$very_active_minutes[myDF$day_of_week == "Friday"])
),
Sat = c(mean(myDF$steps[myDF$day_of_week == "Saturday"]),
mean(myDF$total_miles[myDF$day_of_week == "Saturday"]), mean(myDF$lightly_active_miles[myDF$day_of_week == "Saturday"]),
mean(myDF$moderately_active_miles[myDF$day_of_week == "Saturday"]), mean(myDF$very_active_miles[myDF$day_of_week == "Saturday"]),
mean(myDF$sedentary_minutes[myDF$day_of_week == "Saturday"]), mean(myDF$lightly_active_minutes[myDF$day_of_week == "Saturday"]),
mean(myDF$fairly_active_minutes[myDF$day_of_week == "Saturday"]), mean(myDF$very_active_minutes[myDF$day_of_week == "Saturday"])
),
Sun = c(mean(myDF$steps[myDF$day_of_week == "Sunday"]),
mean(myDF$total_miles[myDF$day_of_week == "Sunday"]), mean(myDF$lightly_active_miles[myDF$day_of_week == "Sunday"]),
mean(myDF$moderately_active_miles[myDF$day_of_week == "Sunday"]), mean(myDF$very_active_miles[myDF$day_of_week == "Sunday"]),
mean(myDF$sedentary_minutes[myDF$day_of_week == "Sunday"]), mean(myDF$lightly_active_minutes[myDF$day_of_week == "Sunday"]),
mean(myDF$fairly_active_minutes[myDF$day_of_week == "Sunday"]), mean(myDF$very_active_minutes[myDF$day_of_week == "Sunday"])
)
)
#sets theme of the dashboard to Flatly
ui<-fluidPage(theme = shinytheme("flatly"),
navbarPage("Fitbit Dashboard",
# tabPanel 1
tabPanel("How to Use",
img(src = "https://assets.phenompeople.com/CareerConnectResources/MERCUS/social/1200x630-1552902977090.jpg", height = 120, width = 220),
h1("Biometric Data Dashboard"),
mainPanel(
strong("Refer to the website here for more information:"),
a("https://www.merck.com/index.html"),
h3("How to Use My Dashboard"),
htmlOutput("text2")
)
),
# tabPanel 2
tabPanel("Average Statistics",
sidebarLayout(
sidebarPanel(br(), br(), br(),
h3("Average Daily Statistics Per Month"),
p("Pick a variable to view the average daily amount of steps,
floors climbed... etc taken each month."),
br(), br(),
selectizeInput("variables", "Variables (Month)", choices = unique(averagesDF$Var), selected = unique(averagesDF$Var)[1]),
br(), br(), br(), br(), br(), br(), br(), br(), br(),
h3("Average Daily Statistics Per Weekday"),
p("Pick a variable to view the average daily amount of steps,
floors climbed... etc taken each week."),
br(), br(),
selectizeInput("variables2", "Variables (Weekdays)", choices = unique(averagesDF2$Var), selected = unique(averagesDF2$Var)[1]),
br(), br(), br(), br()
),
mainPanel(
highchartOutput("plotAvgs"),
highchartOutput("plotAvgs2")
)
)
),
#tabPanel 3
tabPanel("Data Table",
sidebarLayout(
sidebarPanel(
conditionalPanel(
'input.dataset === "myDF"',
checkboxGroupInput("show_vars", "Columns of information to show:",
names(myDF), selected = list("patient_id", "collection_date", "steps"))
),
),
mainPanel(
tabsetPanel(
id = 'dataset',
tabPanel("myDF", DT::dataTableOutput("mytable1"))
)
)
)
)
)
)
server <- function(input, output) {
# You can access the values of the widget (as a vector of Dates)
# with input$dates,
# e.g. output$value <- renderPrint({ input$dates })
# text on tabPanel 1
output$text2 <- renderUI({
#directions for users
HTML(paste("", "1. Average Statistics",
"- Displays the average daily statistics per month or day of week",
"- Monthly Data: select which variable you want to see average statistics for",
"- Weekday Data: select which variable you want to see average statistics for",
"", "2. Data Table ",
"- Select which variables you want visible for each data point",
"- Able to filter through data using filter as well as placing them in order by any variable",
"- Steps and Data are pre chosen variables",
sep="<br/>"))
})
# choose columns to display
myDF2 = myDF[sample(nrow(myDF), nrow(myDF), replace = FALSE), ]
output$mytable1 <- DT::renderDataTable({
DT::datatable(myDF2[, input$show_vars, drop = FALSE], rownames = FALSE, filter = 'top')
})
# sorted columns are colored now because CSS are attached to them
#output$mytable2 <- DT::renderDataTable({
# DT::datatable(finalDataFrame, options = list(orderClasses = TRUE), rownames = FALSE)
#})
#gets the correct variables and stores the data that is required
reactivedf <- reactive({
filtereddf <- averagesDF %>%
dplyr::filter(Var == input$variables)
filtereddf
})
#gets the correct variables and stores the data that is required
reactivedf2 <- reactive({
filtereddf2 <- averagesDF2 %>%
dplyr::filter(Var == input$variables2)
filtereddf2
})
#outputs the object chart that is rendered
output$plotAvgs <- renderHighchart({
highchart() %>%
#creates new columns for each month of the year
hc_add_series(type = "column", reactivedf()$Jan, name = "Jan") %>%
hc_add_series(type = "column", reactivedf()$Feb, name = "Feb") %>%
hc_add_series(type = "column", reactivedf()$Mar, name = "Mar") %>%
hc_add_series(type = "column", reactivedf()$Apr, name = "Apr") %>%
hc_add_series(type = "column", reactivedf()$May, name = "May") %>%
hc_add_series(type = "column", reactivedf()$Jun, name = "Jun") %>%
hc_add_series(type = "column", reactivedf()$Jul, name = "Jul") %>%
hc_add_series(type = "column", reactivedf()$Aug, name = "Aug") %>%
hc_add_series(type = "column", reactivedf()$Sep, name = "Sep") %>%
hc_add_series(type = "column", reactivedf()$Oct, name = "Oct") %>%
hc_add_series(type = "column", reactivedf()$Nov, name = "Nov") %>%
hc_add_series(type = "column", reactivedf()$Dec, name = "Dec") %>%
hc_xAxis(labels = list(enabled = FALSE)) %>%
hc_title(text = input$variables)
})
#outputs the object chart that is rendered
output$plotAvgs2 <- renderHighchart({
highchart() %>%
#creates new columns for each day of the week
hc_add_series(type = "column", reactivedf2()$Mon, name = "Mon") %>%
hc_add_series(type = "column", reactivedf2()$Tue, name = "Tue") %>%
hc_add_series(type = "column", reactivedf2()$Wed, name = "Wed") %>%
hc_add_series(type = "column", reactivedf2()$Thu, name = "Thu") %>%
hc_add_series(type = "column", reactivedf2()$Fri, name = "Fri") %>%
hc_add_series(type = "column", reactivedf2()$Sat, name = "Sat") %>%
hc_add_series(type = "column", reactivedf2()$Sun, name = "Sun") %>%
hc_xAxis(labels = list(enabled = FALSE)) %>%
hc_title(text = input$variables2)
})
}
shinyApp(ui, server)8.17 Critical decisions
Our group decided to use RStudio to test pieces of R code on our own local machines, but we ran into the problem of not being able to collaborate with our team in an efficient way, so we looked for an alternative method where we could all have access to update a file. We decided to create a repository on bitbucket where we all had access to the files and could update everything using git.
After setting up the collaboration, we started working to improve the previous dashboard and implementing new feature. To improve our skills in R shiny dashboard, we each implemented a different form of data visualization, including a histogram, a scatterplot, a box plot, and a pie chart, and implemented them using data from the previous year. Additionally, we looked at code from the previous year’s team and tried to run their app and resolve any issues that occurred.
While implementing the visualization, we faced few difficulties. For instance, in the visualization 1, we have to modify the whole visualization as our initial visualization was having issue to modify itself during the user interaction. Similarly, second visualization was interactive initially and allowed the user to choose which variable they wanted to see on the graph. However, an issue we faced with this was that when the user selected a variable, the values on the y-axis would become N/A and the bars on the graph would all be the same height. After discussing several potential solutions to this issue, we decided that the best option would be to create several static graphs, allowing the user to scroll to the desired graph.
We were originally running into some problems using Highcharter to run the Average Statistics Visualization on some PC's but we found a solution to this by running the dashboard in R version 3.6.3 and restarting R before installing all of the packages.
8.18 Technical Report
During the last sprint, we focused mainly on the visualization and refinement stages of our development process. In the visualization stage, we brainstormed, discussed, and developed the code for the two visualizations we created. In the refinement stage, we reviewed the issues in the code and discussed potential solutions. We then cycled back to the visualization stage in order to implement the solutions we discussed.