PredictionComparison.Rmd

```{r}
library(lubridate)
library(scales)
library(date)
library(dplyr)
library(magrittr)
library(ggplot2)
# work_dir <- "E:/Box Sync/Ambulatory/Neurology/Data from CHP"
work_dir <- "C:/Users/Pranay/Desktop/Hospital Rooms"
#work_dir <- "E:/Box Sync/Ambulatory/Neurology/Data from CHP"
# work_dir <- "C:/Users/Pranay/Desktop/Hospital Rooms"
#setwd(work_dir)

ardata<-read.csv(file.path(work_dir, "ardata.csv"))
colnames(ardata)[which(names(ardata) == "Morning.Afternoon")] <- 'Morning/Afternoon'
colnames(ardata)[which(names(ardata) == "Rented.Rooms.1")] <- 'Rented Rooms'
colnames(ardata)[which(names(ardata) == "Day.of.Week")] <- 'Day of Week'

ardata$Staffed.Rented.Diff<-ardata$Staffed.Rooms - ardata$Rented.Rooms
ardata$Date<- ymd(ardata$Date, tz = "America/New_York")
ardata$`Rented Rooms`<-NULL


```

```{r}

#ardata <- ardata %>%
  #mutate(`Day of Week`=ordered(`Day of Week`, levels= c("Sun", "Mon", 
    #"Tue", "Wed", "Thu", "Fri", "Sat")))
#arbysession <- ardata %>%
  #mutate(releasedrooms = ifelse(Rented.Staffed.Diff >=3, 1, 0),
         #needrooms = ifelse(Staffed.Rented.Diff>=1, 1, 0)) %>%
 #group_by(`Day of Week`, `Morning/Afternoon`) %>%
 #summarize(numsessions=n(),
           #releasedroomsessions=sum(releasedrooms),
           #needroomsessions=sum(needrooms)) %>%
 #mutate(percent.released.sessions=100*releasedroomsessions/numsessions,
        #percent.need.sessions=100*needroomsessions/numsessions)
  
```


Summary by Session:

Steps:
1. Mutate an Indicator Function for every Session based on the differences between Rented and 
  Staffed Rooms-
   Release Room Indicator-For every session which has a difference between Rented and Staffed 
                          Rooms greater than or equal to 3, output 1 for every such session and 
                          0 otherwise.
   Need Room Indicator - For every session which has a difference between Staffed and Rented 
                          Rooms greater than or equal to 2, output 1 for every session and 0 
                          otherwise.
2.Group the dataset by Day of the Week and AM/PM Session.
3.Count the total number of Sessions for every group.
4.Count the total number of Sessions where Rented Rooms need to be released by summing the        entire column having the Release Room Indicator.
5.Count the total number of Sessions where Rented Rooms are needed by summing the entire 
  column having the Need Room Indicator.
6. Calculate the percent of Rooms Released by dividing the Total number of Sesssions where
   Rooms are released by the Total number of Sessions.
7. Calculate the percent of Rooms Needed by dividing the Total number of Sesssions where
   Rooms are needed by the Total number of Sessions.

Group By Day of Week and AM/PM
```{r}
ardata <- ardata %>%
  mutate(`Day of Week`=ordered(`Day of Week`, levels= c("Sun", "Mon", 
    "Tue", "Wed", "Thu", "Fri", "Sat")))
arbysession <- ardata %>%
  mutate(releasedroomsIndicator = ifelse(Rented.Staffed.Diff >=3, 1, 0),
         needroomsIndicator = ifelse(Staffed.Rented.Diff>=2, 1, 0)) %>%
 group_by(`Day of Week`, `Morning/Afternoon`) %>%
 summarize(numsessions=n(),
           ReleaseRoomSessions=sum(releasedroomsIndicator),
           NeedRoomSessions=sum(needroomsIndicator)) %>%
 mutate(Percent.Released.Sessions=round(100*ReleaseRoomSessions/numsessions),
        Percent.Need.Sessions=round(100*NeedRoomSessions/numsessions))

write.table(arbysession, file = "BySess.csv", sep = ",", col.names = NA,
            qmethod = "double")

```

Barplot of rooms needed/released by session
```{r}
arbysession_ug<-ardata %>%
   mutate(releasedroomsIndicator = ifelse(Rented.Staffed.Diff >=3, 1, 0),
         needroomsIndicator = ifelse(Staffed.Rented.Diff>=2, 1, 0)) %>%
  summarize(numsessions=n(),
           ReleaseRoomSessions=sum(releasedroomsIndicator),
           NeedRoomSessions=sum(needroomsIndicator))
  
df <- data.frame(indicate=c( "Release Released", "Rooms Requested"),
                len=c(arbysession_ug$ReleaseRoomSessions,arbysession_ug$NeedRoomSessions))


ggplot(data=df, aes(x=indicate, y=len)) +
  geom_bar(stat="identity")  + scale_y_continuous(breaks= pretty_breaks()) + theme_bw()  + labs(title="Rooms Requested/Released By Total Sessions") + labs(x="Scenario", y="Number of Sessions") 


#ggplot(data=arbysessionroom_release, aes(x=Rented.Staffed.Diff, y=numsessions)) +
  #geom_bar(stat="identity",fill="steelblue") + scale_x_continuous(breaks = 1:20) + scale_y_continuous(breaks= pretty_breaks()) + theme_bw()  + labs(title=" Difference between Rented and Staffed Rooms by Session") + labs(x="Difference = Rented Rooms - Staffed Rooms", y="Number of Sessions") 
  
```


Barplot of Differences between Room by session
```{r}
arbysessionroom_release<-ardata %>%
  filter(Rented.Staffed.Diff>0) %>%
    #mutate(releasedroomsIndicator = ifelse(Rented.Staffed.Diff >=3, 1, 0)) %>%
  group_by(Rented.Staffed.Diff) %>%
   summarize(numsessions=n())
     #ReleaseRoomSessions=sum(releasedroomsIndicator)) #%>%
  #filter(ReleaseRoomSessions>0)
  

arbysessionroom_need<-ardata %>%
  filter(Staffed.Rented.Diff>0) %>%
    #mutate(needroomsIndicator = ifelse(Staffed.Rented.Diff>=2, 1, 0)) %>%
  group_by(Staffed.Rented.Diff) %>%
   summarize(numsessions=n())
     #NeedRoomSessions=sum(needroomsIndicator)) #%>%
  #filter(NeedRoomSessions>0)


ggplot(data=arbysessionroom_release, aes(x=Rented.Staffed.Diff, y=numsessions)) +
  geom_bar(stat="identity",fill="steelblue") + scale_x_continuous(breaks = 1:20) + scale_y_continuous(breaks= pretty_breaks()) + theme_bw()  + labs(title=" Difference between Rented and Staffed Rooms by Total Sessions") + labs(x="Difference = Rented Rooms - Staffed Rooms", y="Number of Sessions") 


ggplot(data=arbysessionroom_need, aes(x=Staffed.Rented.Diff, y=numsessions)) +
  geom_bar(stat="identity",fill="orange") + scale_x_continuous(breaks = 1:20) + scale_y_continuous(breaks= pretty_breaks()) + theme_bw() + labs(title=" Difference between Staffed and Rented Rooms by Total Sessions") + labs(x="Difference = Staffed Rooms - Rented Rooms", y="Number of Sessions") 



#ardata <- ardata %>%
   #mutate(Date=ordered(Date),Date=as.Date(Date))
#mon_am<- ardata %>%
  #filter(`Morning/Afternoon`=="AM",`Day of Week`=="Mon") 
#ggplot(mon_am, aes(Rented.Staffed.Diff)) +
    #geom_histogram(binwidth=1, colour="black", fill="blue") + theme_bw() + scale_x_continuous(breaks = -6:6) + theme_bw() + labs(title=" Difference between Rented and Staffed Rooms on Monday Mornings") +
  #labs(x="Difference = Rented Rooms - Staffed Rooms", y="Number of Sessions") + scale_y_continuous(breaks= pretty_breaks()) +
  #geom_vline(xintercept = 0, linetype="dashed", color = "black", size=1.5) 
  #mon_am$Date<-as.Date(mon_am$Date)
#ggplot(data=mon_am, aes(x=Date)) +                 
  #geom_line(aes(y=Staffed.Rooms, colour="Staffed.Rooms")) +  
  #geom_line(aes(y=Rented.Rooms, colour="Rented.Rooms")) +
  #labs(title="Trend for Rented and Staffed Rooms on Monday Mornings") +
  #labs(x="Date", y="Number of Rooms") + theme_bw() + scale_y_continuous(breaks= 0:14) +
   #scale_colour_manual("", 
                      #breaks = c("Staffed.Rooms", "Rented.Rooms"),
                      #values = c("Staffed.Rooms"="blue", "Rented.Rooms"="violet")) +xlab(" ") + scale_x_date(date_breaks = "1 month",labels = date_format("%b%y")) 





```


Overall Histogram 
```{r}
#ardata <- ardata %>%
   #mutate(Date=ordered(Date),Date=as.Date(Date))
#rent_staff<- ardata %>%
ggplot(ardata, aes(Rented.Staffed.Diff)) +
    geom_histogram(binwidth=1, colour="black", fill="brown") + theme_bw() + scale_x_continuous(breaks = -10:10) + theme_bw() + labs(title=" Difference between Rented and Staffed Rooms") +
  labs(x="Difference = Rented Rooms - Staffed Rooms", y="Number of Sessions") + scale_y_continuous(breaks= pretty_breaks()) +
  geom_vline(xintercept = 0, linetype="dashed", color = "black", size=1.5) 




```








Summary By Rooms:

Procedure:
1.Mutate number of Rented Rooms to release and in need based on the differences between Rented 
  and Staffed Rooms- 
  Released Rooms- If the difference between Rented and Rooms and Staffed Rooms is greater than
                   or equal to 3, Subtract 2 from the difference and the resultant would be the
                   number of Rented Rooms to be released.
  Rooms Needed- If the difference between Staffed Rooms and Rented Rooms is greater than or 
                equal to 2, the number of rooms needed would be the difference minus 1.
2. Group the dataset by Day of Week and AM/PM Session.
3. Count the total number of Staffed and Rented Rooms for every group.
4. Calculate the total number of Rooms to be Release for every group by summing the 'Released
   Rooms' for every group.
5. Calculate the total number of Rooms needed for every group by summing the
   Rooms Needed' for every group.
6.Calculate the percent of the rooms to be released by dividing the Total Number of Released
  Rooms by the Total number of Rented Rooms.
7. Calculate the total number of rooms needed by dividing the Total number of Rooms Needed 
  by the Total number of Rented Rooms. 

```{r}
ardata <- ardata %>%
  mutate(`Day of Week`=ordered(`Day of Week`, levels= c("Sun", "Mon", 
    "Tue", "Wed", "Thu", "Fri", "Sat")))
arbyrooms<- ardata %>%
  mutate(releasedrooms=ifelse(Rented.Staffed.Diff>=3,Rented.Staffed.Diff-2,0),
         needrooms=ifelse(Staffed.Rented.Diff>=2,Staffed.Rented.Diff-1,0)) %>%
 group_by(`Day of Week`, `Morning/Afternoon`) %>%
 summarize(TotalStaffedRooms=sum(Staffed.Rooms),
           TotalRentedRooms=sum(Rented.Rooms),
           TotalRoomsRelease=sum(releasedrooms),
           TotalRoomsNeeded=sum(needrooms)) %>%
 mutate(Percent.Released.Rooms=round(100*TotalRoomsRelease/TotalRentedRooms),
        Percent.Need.Rooms =round(100*TotalRoomsNeeded/TotalRentedRooms),
        Diff.Rented.Requested=TotalRentedRooms-TotalRoomsNeeded)

write.table(arbyrooms, file = "Byroom.csv", sep = ",", col.names = NA,
            qmethod = "double")

```








1. Monday: Mornings and Afternoons

Plot 1: Histogram distribution showing difference between Rented and Staffed Rooms on Monday            Mornings.
Plot 2: Trend for Staffed Rooms and Rented Rooms on Monday Mornings.
Plot 3:Histogram distribution showing difference between Rented and Staffed Rooms on Monday            Afternoons.
Plot 4:Trend for Staffed Rooms and Rented Rooms on Monday Afternoons.
```{r}
#Morning
ardata <- ardata %>%
   mutate(Date=ordered(Date),Date=as.Date(Date))
mon_am<- ardata %>%
  filter(`Morning/Afternoon`=="AM",`Day of Week`=="Mon") 
ggplot(mon_am, aes(Rented.Staffed.Diff)) +
    geom_histogram(binwidth=1, colour="black", fill="blue") + theme_bw() + scale_x_continuous(breaks = -6:6) + theme_bw() + labs(title=" Difference between Rented and Staffed Rooms on Monday Mornings") +
  labs(x="Difference = Rented Rooms - Staffed Rooms", y="Number of Sessions") + scale_y_continuous(breaks= pretty_breaks()) +
  geom_vline(xintercept = 0, linetype="dashed", color = "black", size=1.5) 
  #mon_am$Date<-as.Date(mon_am$Date)
ggplot(data=mon_am, aes(x=Date)) +                 
  geom_line(aes(y=Staffed.Rooms, colour="Staffed.Rooms")) +  
  geom_line(aes(y=Rented.Rooms, colour="Rented.Rooms")) +
  labs(title="Trend for Rented and Staffed Rooms on Monday Mornings") +
  labs(x="Date", y="Number of Rooms") + theme_bw() + scale_y_continuous(breaks= 0:14) +
   scale_colour_manual("", 
                      breaks = c("Staffed.Rooms", "Rented.Rooms"),
                      values = c("Staffed.Rooms"="blue", "Rented.Rooms"="violet")) +xlab(" ") + scale_x_date(date_breaks = "1 month",labels = date_format("%b%y")) 


#Afternoon
mon_pm<- ardata %>%
  filter(`Morning/Afternoon`=="PM",`Day of Week`=="Mon") 
ggplot(mon_pm, aes(x=Rented.Staffed.Diff)) +
    geom_histogram(binwidth=1, colour="black", fill="red") + theme_bw() + scale_x_continuous(breaks = -6:6) + theme_bw() + labs(title=" Difference between Rented and Staffed Rooms on Monday Afternoons") +
  labs(x="Difference = Rented Rooms - Staffed Rooms", y="Number of Sessions") + scale_y_continuous(breaks= pretty_breaks()) + 
  geom_vline(xintercept = 0, linetype="dashed", color = "black", size=1.5)
ggplot(data=mon_pm, aes(x=Date)) +                 
  geom_line(aes(y=Staffed.Rooms, colour="Staffed.Rooms")) +  
  geom_line(aes(y=Rented.Rooms, colour="Rented.Rooms")) +
  labs(title="Trend for Rented and Staffed Rooms on Monday Afternoons") +
  labs(x="Date", y="Number of Rooms") + theme_bw() + scale_y_continuous(breaks= 0:10) +
   scale_colour_manual("", 
                      breaks = c("Staffed.Rooms", "Rented.Rooms"),
                      values = c("Staffed.Rooms"="red", "Rented.Rooms"="green")) +xlab(" ")  + scale_x_date(date_breaks = "1 month",labels = date_format("%b%y"))


```

2. Tuesday: Mornings and Afternoons

Plot 1: Histogram distribution showing difference between Rented and Staffed Rooms on Tuesday           Mornings.
Plot 2: Trend for Staffed Rooms and Rented Rooms on Tuesday Mornings.
Plot 3:Histogram distribution showing difference between Rented and Staffed Rooms on Tuesday            Afternoons.
Plot 4:Trend for Staffed Rooms and Rented Rooms on Tuesday Afternoons.
```{r}
#Morning
tue_am<- ardata %>%
  filter(`Morning/Afternoon`=="AM",`Day of Week`=="Tue") 
ggplot(tue_am, aes(x=Rented.Staffed.Diff)) +
    geom_histogram(binwidth=1, colour="black", fill="blue") + theme_bw() + scale_x_continuous(breaks = -8:8) + theme_bw() + labs(title=" Difference between Rented and Staffed Rooms on Tuesday Mornings") +
  labs(x="Difference = Rented Rooms - Staffed Rooms", y="Number of Sessions") + scale_y_continuous(breaks= pretty_breaks()) + 
  geom_vline(xintercept = 0, linetype="dashed", color = "black", size=1.5)
ggplot(data=tue_am, aes(x=Date)) +                 
  geom_line(aes(y=Staffed.Rooms, colour="Staffed.Rooms")) +  
  geom_line(aes(y=Rented.Rooms, colour="Rented.Rooms")) +
  labs(title="Trend for Rented and Staffed Rooms on Tuesday Mornings") +
  labs(x="Date", y="Number of Rooms") + theme_bw() + scale_y_continuous(breaks= 0:20) +
   scale_colour_manual("", 
                      breaks = c("Staffed.Rooms", "Rented.Rooms"),
                      values = c("Staffed.Rooms"="blue", "Rented.Rooms"="violet")) +xlab(" ")  + scale_x_date(date_breaks = "1 month",labels = date_format("%b%y"))



#Afternoon
tue_pm<- ardata %>%
  filter(`Morning/Afternoon`=="PM",`Day of Week`=="Tue") 
ggplot(tue_pm, aes(x=Rented.Staffed.Diff)) +
    geom_histogram(binwidth=1, colour="black", fill="red") + theme_bw() + scale_x_continuous(breaks = -6:6) + theme_bw() + labs(title=" Difference between Rented and Staffed Rooms on Tuesday Afternoons") +
  labs(x="Difference = Rented Rooms - Staffed Rooms", y="Number of Sessions") + scale_y_continuous(breaks= pretty_breaks()) + 
  geom_vline(xintercept = 0, linetype="dashed", color = "black", size=1.5)
#tue_pm$Date<-as.Date(tue_pm$Date)
ggplot(data=tue_pm, aes(x=Date)) +                 
  geom_line(aes(y=Staffed.Rooms, colour="Staffed.Rooms")) +  
  geom_line(aes(y=Rented.Rooms, colour="Rented.Rooms")) +
  labs(title="Trend for Rented and Staffed Rooms on Tuesday Afternoons") +
  labs(x="Date", y="Number of Rooms") + theme_bw() + scale_y_continuous(breaks= 0:12) +
   scale_colour_manual("", 
                      breaks = c("Staffed.Rooms", "Rented.Rooms"),
                      values = c("Staffed.Rooms"="red", "Rented.Rooms"="green")) +xlab(" ")  + scale_x_date(date_breaks = "1 month",labels = date_format("%b%y"))


```

3. Wednesday: Mornings and Afternoons

Plot 1: Histogram distribution showing difference between Rented and Staffed Rooms on Wednesday            Mornings.
Plot 2: Trend for Staffed Rooms and Rented Rooms on Wednesday Mornings.
Plot 3:Histogram distribution showing difference between Rented and Staffed Rooms on Wednesday            Afternoons.
Plot 4:Trend for Staffed Rooms and Rented Rooms on Wednesday Afternoons.
```{r}
#Morning
wed_am<- ardata %>%
  filter(`Morning/Afternoon`=="AM",`Day of Week`=="Wed") 
ggplot(wed_am, aes(x=Rented.Staffed.Diff)) +
    geom_histogram(binwidth=1, colour="black", fill="blue") + theme_bw() + scale_x_continuous(breaks = -6:6) + theme_bw() + labs(title=" Difference between Rented and Staffed Rooms on Wednesday Mornings") +
  labs(x="Difference = Rented Rooms - Staffed Rooms", y="Number of Sessions") + scale_y_continuous(breaks= pretty_breaks()) + 
  geom_vline(xintercept = 0, linetype="dashed", color = "black", size=1.5)
ggplot(data=wed_am, aes(x=Date)) +                 
  geom_line(aes(y=Staffed.Rooms, colour="Staffed.Rooms")) +  
  geom_line(aes(y=Rented.Rooms, colour="Rented.Rooms")) +
  labs(title="Trend for Rented and Staffed Rooms on Wednesday Mornings") +
  labs(x="Date", y="Number of Rooms") + theme_bw() + scale_y_continuous(breaks= 0:10) +
   scale_colour_manual("", 
                      breaks = c("Staffed.Rooms", "Rented.Rooms"),
                      values = c("Staffed.Rooms"="blue", "Rented.Rooms"="violet")) +xlab(" ") + scale_x_date(date_breaks = "1 month",labels = date_format("%b%y"))


#Afternoon
wed_pm<- ardata %>%
  filter(`Morning/Afternoon`=="PM",`Day of Week`=="Wed") 
ggplot(wed_pm, aes(x=Rented.Staffed.Diff)) +
    geom_histogram(binwidth=1, colour="black", fill="red") + theme_bw() + scale_x_continuous(breaks = -6:6) + theme_bw() + labs(title=" Difference between Rented and Staffed Rooms on Wednesday Afternoons") +
  labs(x="Difference = Rented Rooms - Staffed Rooms", y="Number of Sessions") + scale_y_continuous(breaks= pretty_breaks()) + 
  geom_vline(xintercept = 0, linetype="dashed", color = "black", size=1.5)
ggplot(data=wed_pm, aes(x=Date)) +                 
  geom_line(aes(y=Staffed.Rooms, colour="Staffed.Rooms")) +  
  geom_line(aes(y=Rented.Rooms, colour="Rented.Rooms")) +
  labs(title="Trend for Rented and Staffed Rooms on Wednesday Afternoons") +
  labs(x="Date", y="Number of Rooms") + theme_bw() + scale_y_continuous(breaks= 0:10) +
   scale_colour_manual("", 
                      breaks = c("Staffed.Rooms", "Rented.Rooms"),
                      values = c("Staffed.Rooms"="red", "Rented.Rooms"="green")) +xlab(" ")  + scale_x_date(date_breaks = "1 month",labels = date_format("%b%y"))


```

<<<<<<< HEAD
4. Thursday: Mornings and Afternoons

Plot 1: Histogram distribution showing difference between Rented and Staffed Rooms on Thursday            Mornings.
Plot 2: Trend for Staffed Rooms and Rented Rooms on Thursday Mornings.
Plot 3:Histogram distribution showing difference between Rented and Staffed Rooms on Thursday            Afternoons.
Plot 4:Trend for Staffed Rooms and Rented Rooms on Thursday Afternoons.
```{r}
#Morning
thur_am<- ardata %>%
  filter(`Morning/Afternoon`=="AM",`Day of Week`=="Thu") 
ggplot(thur_am, aes(x=Rented.Staffed.Diff)) +
    geom_histogram(binwidth=1, colour="black", fill="blue") + theme_bw() + scale_x_continuous(breaks = -12:7) + theme_bw() + labs(title=" Difference between Rented and Staffed Rooms on Thursday Mornings") +
  labs(x="Difference = Rented Rooms - Staffed Rooms", y="Number of Sessions") + scale_y_continuous(breaks= pretty_breaks()) + 
  geom_vline(xintercept = 0, linetype="dashed", color = "black", size=1.5)
ggplot(data=thur_am, aes(x=Date)) +                 
  geom_line(aes(y=Staffed.Rooms, colour="Staffed.Rooms")) +  
  geom_line(aes(y=Rented.Rooms, colour="Rented.Rooms")) +
  labs(title="Trend for Rented and Staffed Rooms on Thursday Mornings") +
  labs(x="Date", y="Number of Rooms") + theme_bw() + scale_y_continuous(breaks= 0:16) +
   scale_colour_manual("", 
                      breaks = c("Staffed.Rooms", "Rented.Rooms"),
                      values = c("Staffed.Rooms"="blue", "Rented.Rooms"="violet")) +xlab(" ")  + scale_x_date(date_breaks = "1 month",labels = date_format("%b%y"))


#AFternoon
thur_pm<- ardata %>%
  filter(`Morning/Afternoon`=="PM",`Day of Week`=="Thu") 
ggplot(thur_pm, aes(x=Rented.Staffed.Diff)) +
    geom_histogram(binwidth=1, colour="black", fill="red") + theme_bw() + scale_x_continuous(breaks = -3:7) + theme_bw() + labs(title=" Difference between Rented and Staffed Rooms on Thursday Afternoons") +
  labs(x="Difference = Rented Rooms - Staffed Rooms", y="Number of Sessions") + scale_y_continuous(breaks= pretty_breaks()) +
  geom_vline(xintercept = 0, linetype="dashed", color = "black", size=1.5) 
ggplot(data=thur_pm, aes(x=Date)) +                 
  geom_line(aes(y=Staffed.Rooms, colour="Staffed.Rooms")) +  
  geom_line(aes(y=Rented.Rooms, colour="Rented.Rooms")) +
  labs(title="Trend for Rented and Staffed Rooms on Thursday Afternoons") +
  labs(x="Date", y="Number of Rooms") + theme_bw() + scale_y_continuous(breaks= 0:10) +
   scale_colour_manual("", 
                      breaks = c("Staffed.Rooms", "Rented.Rooms"),
                      values = c("Staffed.Rooms"="red", "Rented.Rooms"="green")) +xlab(" ")  + scale_x_date(date_breaks = "1 month",labels = date_format("%b%y"))


```

5. Friday: Mornings and Afternoons

Plot 1: Histogram distribution showing difference between Rented and Staffed Rooms on Friday            Mornings.
Plot 2: Trend for Staffed Rooms and Rented Rooms on Friday Mornings.
Plot 3:Histogram distribution showing difference between Rented and Staffed Rooms on Friday            Afternoons.
Plot 4:Trend for Staffed Rooms and Rented Rooms on Friday Afternoons.
```{r}
#Friday Morning
fri_am<- ardata %>%
  filter(`Morning/Afternoon`=="AM",`Day of Week`=="Fri")
ggplot(fri_am, aes(x=Rented.Staffed.Diff)) +
    geom_histogram(binwidth=1, colour="black", fill="blue") + theme_bw() + scale_x_continuous(breaks = -7:11) + theme_bw() + labs(title=" Difference between Rented and Staffed Rooms on Friday Mornings") +
  labs(x="Difference = Rented Rooms - Staffed Rooms", y="Number of Sessions") + scale_y_continuous(breaks= pretty_breaks()) +
  geom_vline(xintercept = 0, linetype="dashed", color = "black", size=1.5)
ggplot(data=fri_am, aes(x=Date)) +                 
  geom_line(aes(y=Staffed.Rooms, colour="Staffed.Rooms")) +  
  geom_line(aes(y=Rented.Rooms, colour="Rented.Rooms")) +
  labs(title="Trend for Rented and Staffed Rooms on Friday Mornings") +
  labs(x="Date", y="Number of Rooms") + theme_bw() + scale_y_continuous(breaks=0:18) +
   scale_colour_manual("", 
                      breaks = c("Staffed.Rooms", "Rented.Rooms"),
                      values = c("Staffed.Rooms"="blue", "Rented.Rooms"="violet")) +xlab(" ")  + scale_x_date(date_breaks = "1 month",labels = date_format("%b%y"))

#Friday Afternoon
fri_pm<- ardata %>%
  filter(`Morning/Afternoon`=="PM",`Day of Week`=="Fri")
ggplot(fri_pm, aes(x=Rented.Staffed.Diff)) +
    geom_histogram(binwidth=1, colour="black", fill="red") + theme_bw() + scale_x_continuous(breaks = -3:8) + theme_bw() + labs(title=" Difference between Rented and Staffed Rooms on Friday Afternoons") +
  labs(x="Difference = Rented Rooms - Staffed Rooms", y="Number of Sessions") + scale_y_continuous(breaks= pretty_breaks()) +
  geom_vline(xintercept = 0, linetype="dashed", color = "black", size=1.5)
ggplot(data=fri_pm, aes(x=Date)) +                 
  geom_line(aes(y=Staffed.Rooms, colour="Staffed.Rooms")) +  
  geom_line(aes(y=Rented.Rooms, colour="Rented.Rooms")) +
  labs(title="Trend for Rented and Staffed Rooms on Friday Afternoons") +
  labs(x="Date", y="Number of Rooms") + theme_bw() + scale_y_continuous(breaks= 0:12) +
   scale_colour_manual("", 
                      breaks = c("Staffed.Rooms", "Rented.Rooms"),
                      values = c("Staffed.Rooms"="red", "Rented.Rooms"="green")) +xlab(" ")  + scale_x_date(date_breaks = "1 month",labels = date_format("%b%y"))



```