---

title: "topChef"

output: github_document

---

```{r, include = FALSE, message=FALSE}

knitr::opts_chunk$set(

collapse = TRUE,

comment = "#>",

warning=FALSE

)

```

## 1. Introduction to topChef

topChef is a collection of data sets detailing events across all seasons of Top Chef US and Top Chef Masters US and one season of Top Chef Canada. It includes Chef information, challenge descriptions, challenge winners, episode information, guest judge names, and reward/prize information.

## 2. Installation

Not yet on CRAN. So please use: *devtools::install.packages("celevitz/topChef")*.

If it's not appearing to be updated, restart your R sessions, install it again, and call it into your library.

```{r InstallPackage , message=FALSE,warning=FALSE}

devtools::install_github("celevitz/topChef")

```

## 3. References & Acknowlegements

Data were collected manually while watching each season of Top Chef. Additional data were collected from <https://en.wikipedia.org/wiki/Top_Chef>. My Top Chef data journey was inspired by <https://topchefstats.com/>.

Huge thanks to <https://github.com/doehm> for all his support!

```{r Library set up, message=FALSE, warning=FALSE,echo=FALSE,eval=TRUE}

library(topChef)

library(tidyr)

library(dplyr)

```

## 4. Overview of datasets

Across datasets, key joining variables include:

* `chef`

* `szn`

* `sznnumber`

* `series`

* `episode`

### 4.a. Datasets

#### 4.a.i. Chef details

A tibble containing information about Chefs for each season they are in, including placement and gender. For some but not all seasons, there is also information on hometown, current city of residence, age, a flag for whether they are a person of color, and their occupation.

```{r Chef details table , eval=TRUE}

chefdetails

```

#### 4.a.ii. Challenge descriptions

A tibble containing information about each challenge that the Chefs compete in.

```{r Challenge descriptions table , eval=TRUE}

challengedescriptions

```

#### 4.a.iii. Challenge wins

A tibble containing win and loss data for each chef in each episode.

```{r Challenge results table , eval=TRUE}

challengewins

```

#### 4.a.iv. Judges

A tibble containing information about who were the guest judges for each challenge.

```{r Guest Judges table , eval=TRUE}

judges

```

#### 4.a.v. Rewards

A tibble containing information about rewards and prizes won by challenge.

```{r Rewards table , eval=TRUE}

rewards

```

#### 4.a.vi. Episode info

A tibble containing information about each episode.

```{r Episode information table , eval=TRUE}

episodeinfo

```

### 4.b. Example using multiple datasets: How many elimination challenge wins did Top Chef winners have?

##### 4.b.i. Visualization

```{r EliminationWinsForTopChefsVisualization, eval=TRUE,echo=FALSE,message=FALSE}

library(ggplot2)

chefdetails %>%

# Keep just winners and relevant variables (just the winning chef's name)

filter(placement == 1) %>%

select(chef, gender,series,szn) %>%

# Bring on the challenge result data

left_join(challengewins) %>%

# Keep just elimination & sudden death quickfire challenges

filter(challenge_type %in% c("Elimination","Quickfire Elimination","Sudden Death Quickfire")) %>%

# Get the number of wins for each person

group_by(series,szn,sznnumber,chef,gender) %>%

filter(outcome %in% c("WIN","WINNER")) %>%

summarise(wins=n()) %>%

# Plot

ggplot(aes(x=wins,y=chef,color=gender,fill=gender,shape=series)) +

geom_point(size=4) +

theme_minimal() +

# beautification

scale_color_manual(values=c("#1170AA", "#A3ACB9")) +

scale_shape_manual(values=c(0,17,1)) +

scale_x_continuous(name="Number of elimination wins",

breaks=seq(1,9,1),

limits=c(.8,9.5),

labels=seq(1,9,1)) +

theme(panel.grid = element_blank(),

title=element_text(size=12),

axis.line.x=element_line(color="black"),

axis.ticks.x=element_line(color="black"),

axis.text.x=element_text(size=12),

axis.title.x=element_text(size=12)) +

labs(title="Number of elimination wins of Top Chefs"

,caption="I should have ordered by # of wins...")

```

##### 4.b.ii. Code

```{r EliminationWinsForTopChefsCode, eval=FALSE,echo=TRUE,message=FALSE}

library(ggplot2)

chefdetails %>%

# Keep just winners and relevant variables (just the winning chef's name)

filter(placement == 1) %>%

select(chef, gender,series,szn) %>%

# Bring on the challenge result data

left_join(challengewins) %>%

# Keep just elimination & sudden death quickfire challenges

filter(challenge_type %in% c("Elimination","Quickfire Elimination","Sudden Death Quickfire")) %>%

# Get the number of wins for each person

group_by(series,szn,sznnumber,chef,gender) %>%

filter(outcome %in% c("WIN","WINNER")) %>%

summarise(wins=n()) %>%

# Plot

ggplot(aes(x=wins,y=chef,color=gender,fill=gender,shape=series)) +

geom_point(size=4) +

theme_minimal() +

# beautification

scale_color_manual(values=c("#1170AA", "#A3ACB9")) +

scale_shape_manual(values=c(0,17,1)) +

scale_x_continuous(name="Number of elimination wins",

breaks=seq(1,9,1),

limits=c(.8,9.5),

labels=seq(1,9,1)) +

theme(panel.grid = element_blank(),

title=element_text(size=12),

axis.line.x=element_line(color="black"),

axis.ticks.x=element_line(color="black"),

axis.text.x=element_text(size=12),

axis.title.x=element_text(size=12)) +

labs(title="Number of elimination wins of Top Chefs"

,caption="I should have ordered by # of wins...")

```

### 4.c. Example using multiple datasets: winning an episode

#### 4.c.i. Visualization

```{r WonEpisodesViz, eval=TRUE,echo=FALSE,message=FALSE}

library(ggplot2)

library(topChef)

## Won the episode

# Data set up

# how many episodes each chef won in each season

wonepi <- challengewins %>% select(!rating) %>%

mutate(challenge_type=case_when(challenge_type %in% c("Quickfire Elimination","Sudden Death Quickfire") ~ "Quickfire"

,TRUE ~ challenge_type)) %>%

filter(outcome %in% c("WIN","WINNER") & challenge_type %in% c("Elimination","Quickfire")) %>%

distinct() %>%

pivot_wider(names_from=challenge_type,values_from=outcome) %>%

filter(Quickfire == "WIN" & (Elimination %in% c("WIN","WINNER"))) %>%

# were there multiple winners

group_by(series,szn,sznnumber, episode) %>%

mutate(nchefs=n()

,chef=ifelse(nchefs > 1,paste0(chef,"*"),chef)) %>%

# how many times did each chef win an episode?

ungroup() %>% group_by(series,szn,sznnumber,chef) %>%

summarise(n=n()) %>%

select(series,szn,sznnumber,chef,n) %>%

distinct() %>%

mutate(sznnumberchar=case_when(sznnumber <= 9 ~paste0("0",as.character(sznnumber))

,TRUE ~as.character(sznnumber)) ) %>%

filter(series == "US")

# visualize

wonepivizdata <- wonepi %>%

# add on placement of chefs, just for those who have won an episode

left_join(topChef::chefdetails %>% select(szn,sznnumber,chef,placement)) %>%

mutate(placement=as.character(case_when(placement > 5 ~ 5

,TRUE ~ placement))

,placement = case_when(placement == "5" ~ "5th or lower"

,placement == "1" ~ "1st"

,placement == "2" ~ "2nd"

,placement == "3" ~ "3rd"

,placement == "4" ~ "4th"

,TRUE ~ placement))

wonepivizdata <- wonepivizdata[order(wonepivizdata$sznnumberchar,wonepivizdata$n,wonepivizdata$chef),]

## Visualization

wonepivizdata %>%

ggplot(aes(x=sznnumberchar,y=n,label=chef,fill=factor(placement))) +

scale_fill_manual(values=c("#141B41","#1170AA","#5fa2ce","#a3cce9","#ababab")) +

geom_col(position="stack",color="white") +

geom_text(size = 1, position = position_stack(vjust=.5),angle=15

,color=case_when(wonepivizdata$placement == "1st"~ "white"

,wonepivizdata$placement == "2nd" ~ "white"

,TRUE ~ "black"

,is.na(wonepivizdata$placement) ~ "black")) +

labs(title="A Chef has won an episode in all but Seasons 1 and 16"

,subtitle="Winning an episode means that one chef won both the Quickfire and Elimination Challenges.\nThis has happened 35 times. Eight Top Chefs won an episode in their season.\n

Brooke is the only one to do it in two different seasons. In all but three seasons, only

one or two chefs won an episode. In seasons four and seven, three chefs won an

episode. In season 11, six chefs won an episode. Stefan and Gregory G. are the only

chefs to have ever won three episodes in one season. Four chefs have won two

episodes in one season: Angelo S., Ed C., Tiffany D., and Brooke W.

* Travis and Carlos won the team quickfire and the paired elimination challenge."

,caption= "Data: github.com/celevitz/topChef /// Twitter@carlylevitz") +

ylab("\n# of times a chef won an episode this season\n") +

xlab("Season") +

theme_minimal() +

theme(panel.grid = element_blank()

,axis.line.x=element_line(color="black")

,axis.ticks.x=element_line(color="black")

,axis.line.y=element_line(color="black")

,axis.ticks.y=element_line(color="black")

,axis.text.x=element_text(color="black")

,axis.text.y=element_text(color="black")

,axis.title.x = element_text(color="black")

,axis.title.y = element_text(color="black") ) +

guides(fill = guide_legend(title = "Placement"))

```

#### 4.c.ii. Code

```{r WonEpisodesViz_Code, eval=FALSE,echo=TRUE,message=FALSE}

library(ggplot2)

library(topChef)

## Won the episode

# Data set up

# how many episodes each chef won in each season

wonepi <- challengewins %>% select(!rating) %>%

mutate(challenge_type=case_when(challenge_type %in% c("Quickfire Elimination","Sudden Death Quickfire") ~ "Quickfire"

,TRUE ~ challenge_type)) %>%

filter(outcome %in% c("WIN","WINNER") & challenge_type %in% c("Elimination","Quickfire")) %>%

distinct() %>%

pivot_wider(names_from=challenge_type,values_from=outcome) %>%

filter(Quickfire == "WIN" & (Elimination %in% c("WIN","WINNER"))) %>%

# were there multiple winners

group_by(series,szn,sznnumber, episode) %>%

mutate(nchefs=n()

,chef=ifelse(nchefs > 1,paste0(chef,"*"),chef)) %>%

# how many times did each chef win an episode?

ungroup() %>% group_by(series,szn,sznnumber,chef) %>%

summarise(n=n()) %>%

select(series,szn,sznnumber,chef,n) %>%

distinct() %>%

mutate(sznnumberchar=case_when(sznnumber <= 9 ~paste0("0",as.character(sznnumber))

,TRUE ~as.character(sznnumber)) ) %>%

filter(series == "US")

# visualize

wonepivizdata <- wonepi %>%

# add on placement of chefs, just for those who have won an episode

left_join(topChef::chefdetails %>% select(szn,sznnumber,chef,placement)) %>%

mutate(placement=as.character(case_when(placement > 5 ~ 5

,TRUE ~ placement))

,placement = case_when(placement == "5" ~ "5th or lower"

,placement == "1" ~ "1st"

,placement == "2" ~ "2nd"

,placement == "3" ~ "3rd"

,placement == "4" ~ "4th"

,TRUE ~ placement))

wonepivizdata <- wonepivizdata[order(wonepivizdata$sznnumberchar,wonepivizdata$n,wonepivizdata$chef),]

## Visualization

wonepivizdata %>%

ggplot(aes(x=sznnumberchar,y=n,label=chef,fill=factor(placement))) +

scale_fill_manual(values=c("#141B41","#1170AA","#5fa2ce","#a3cce9","#ababab")) +

geom_col(position="stack",color="white") +

geom_text(size = 1, position = position_stack(vjust=.5),angle=15

,color=case_when(wonepivizdata$placement == "1st"~ "white"

,wonepivizdata$placement == "2nd" ~ "white"

,TRUE ~ "black"

,is.na(wonepivizdata$placement) ~ "black")) +

labs(title="A Chef has won an episode in all but Seasons 1 and 16"

,subtitle="Winning an episode means that one chef won both the Quickfire and Elimination Challenges.\nThis has happened 35 times. Eight Top Chefs won an episode in their season.\n

Brooke is the only one to do it in two different seasons. In all but three seasons, only

one or two chefs won an episode. In seasons four and seven, three chefs won an

episode. In season 11, six chefs won an episode. Stefan and Gregory G. are the only

chefs to have ever won three episodes in one season. Four chefs have won two

episodes in one season: Angelo S., Ed C., Tiffany D., and Brooke W.

* Travis and Carlos won the team quickfire and the paired elimination challenge."

,caption= "Data: github.com/celevitz/topChef /// Twitter@carlylevitz") +

ylab("\n# of times a chef won an episode this season\n") +

xlab("Season") +

theme_minimal() +

theme(panel.grid = element_blank()

,axis.line.x=element_line(color="black")

,axis.ticks.x=element_line(color="black")

,axis.line.y=element_line(color="black")

,axis.ticks.y=element_line(color="black")

,axis.text.x=element_text(color="black")

,axis.text.y=element_text(color="black")

,axis.title.x = element_text(color="black")

,axis.title.y = element_text(color="black") ) +

guides(fill = guide_legend(title = "Placement"))

```

## 5. Weighted Index Function

I created a weighted index to compare chefs within and across seasons. I am still working on it, but this is how things currently stand: The weighted index is calculated by assigning a certain number of points to different outcomes. For the sake of simplicity, Sudden Death Quickfires and Quickfire Elimination Challenges are counted as Elimination Challenges. The scoring is as follows:

* Elimination win = 7 points

* Elimination high = 3 points

* Elimination low = -3 points

* Eliminated = -7 points

* Quickfire win = 4 points

* Quickfire high = 2 points

* Quickfire low = -2 points

The `weightedindex` function takes the following parameters:

* `series`: US, US Masters, or Canada

* `seasonnumber`: Values between 1 and 20 for Top Chef US; 1 through 5 for US Masters; and 6 for Canada

* `numberofelimchalls`: Number of elimination challenges through which you want to calculate the index. Values between 1 and 20.

* `numberofquickfires`: Number of quickfire challenges through which you want to calculate the index. Values between 1 and 20.

Currently, the function will include more Quickfire challenges than you have specified in the function. It will keep the episodes in each season through the episode with the Nth Elimination Challenge or the Nth Quickfire, whichever is the higher episode. This was a way to hold constant the number of challenges that have occurred so that chefs could be more comparable across seasons.

The `weightedindex` function exports a tibble with the variables of: series, season name, season number, chef, chef's placement, number of elimination challenges wins/highs/lows/outs, number of quickfire challenge wins/highs/lows, and the weighted index score.

### 5.a. Distribution of Weighted Index Scores at the Completion of All Seasons

```{r Viz_DensityPlot_SetUp, eval=TRUE,echo=FALSE,message=FALSE,include=FALSE}

library(topChef)

library(tidyr)

library(dplyr)

library(RColorBrewer)

## Get the index for all seasons, all episodes

allseasons <- weightedindex("US",1,20,20)

for (season in seq(2,20,1)) {

allseasons <- rbind(allseasons,weightedindex("US",season,20,20))

}

# for sorting reasons, have the season be a character

allseasons$seasonnumchar[allseasons$sznnumber <= 9] <-

paste0("0",as.character(allseasons$sznnumber[allseasons$sznnumber <= 9]))

allseasons$seasonnumchar[allseasons$sznnumber > 9] <-

as.character(allseasons$sznnumber[allseasons$sznnumber > 9])

# Standard deviations

standarddevs <-

aggregate(allseasons$indexWeight,by=list(allseasons$szn,allseasons$sznnumber),FUN=sd)

names(standarddevs) <- c("szn","sznnumber","sd")

standarddevs <- standarddevs[order(standarddevs$sd),] %>%

left_join(topChef::chefdetails %>%

filter(placement == 1) %>%

select(szn,sznnumber,chef)) %>%

mutate(y=0

,labeljust=case_when(szn %in% c("Chicago","Los Angeles","Portland","Seattle") ~ 1

,TRUE ~ 0))

```

```{r Viz_DensityPlot, eval=TRUE,echo=FALSE,message=FALSE}

allseasons %>%

ggplot(aes(x=indexWeight,color=szn,fill=szn)) +

geom_density(alpha=.01) +

labs(title="Density Plot of Weighted Index Score for Every Top Chef US Season") +

scale_color_manual(values=c("#A6CEE3", "#1F78B4", "#B2DF8A", "#33A02C"

,"#FB9A99", "#E31A1C", "#FDBF6F", "#FF7F00"

,"#CAB2D6", "#6A3D9A","#FFFF99","#B15928"

,"#1B9E77", "#D95F02" ,"#7570B3" ,"#E7298A"

,"#66A61E", "#E6AB02", "#A6761D", "#666666")) +

scale_fill_manual(values=c("#A6CEE3", "#1F78B4", "#B2DF8A", "#33A02C"

,"#FB9A99", "#E31A1C", "#FDBF6F", "#FF7F00"

,"#CAB2D6", "#6A3D9A","#FFFF99","#B15928"

,"#1B9E77", "#D95F02" ,"#7570B3" ,"#E7298A"

,"#66A61E", "#E6AB02", "#A6761D", "#666666")) +

theme_minimal()+

theme(legend.position="right"

,legend.key.size = unit(.5, 'cm')) +

guides(color=guide_legend(title="Season name")

,fill=guide_legend(title="Season name")) +

ylab("Density") + xlab("Weighted Index Score")

allseasons %>%

ggplot(aes(x=seasonnumchar,y=indexWeight) ) +

geom_boxplot() +

# add horizontal line at 0

geom_rect(aes(xmin=0

,xmax=20

,ymin=-.5

,ymax=0.5)

,fill="#ffbc69") +

theme_minimal() +

labs(title=paste0("Top Chef Weighted Index: Comparing All Completed Seasons")

,subtitle="Box plots show the minimum, first quartile, median, third quartile, and

maximum weighted index scores for each season. Circles indicate outliers.")+

scale_x_discrete(labels=unique(allseasons$szn[order(allseasons$sznnumber)])) +

theme_minimal() +

ylab("Index score") + xlab("") +

theme(panel.grid = element_blank()

,axis.text.x=element_text(angle=90)

)

standarddevs %>%

ggplot(aes(x=sd,y=y,label=szn)) +

geom_hline(yintercept=0, color="darkcyan") +

geom_point(alpha=.3,size=4) +

theme_minimal() +

ylab("") + xlab("Standard deviation of weighted index scores of each season") +

theme(panel.grid = element_blank()

,axis.line.x=element_line(color="black")

,axis.ticks.x=element_line(color="black")

,axis.line.y=element_blank()

,axis.text.y=element_blank()) +

labs(title="Standard deviation of each Top Chef season's weighted index scores"

,subtitle="Lower standard deviations indicate more even distribution of scores in that season") +

geom_text(angle=45,hjust=standarddevs$labeljust) +

scale_x_continuous(lim=c(7.5,23),breaks=seq(8,22,2),labels=seq(8,22,2))

```

### 5.b. Examples that use the `weightedindex` function: full seasons

##### 5.b.i. Visualizations

```{r Viz_FinalIndex_Names_Setup, eval=TRUE,echo=FALSE,message=FALSE,include=FALSE,warning=FALSE}

library(topChef)

library(ggplot2)

library(tidyr)

library(dplyr)

## Get the index for all seasons

allseasons <- weightedindex("US",1,20,20)

for (season in seq(2,20,1)) {

allseasons <- rbind(allseasons,weightedindex("US",season,20,20))

}

```

```{r Viz_FinalIndex_Names, eval=TRUE,echo=FALSE,message=FALSE,warning=FALSE}

## Graph it

# for sorting reasons, have the season be a character

allseasons$seasonnumchar[allseasons$sznnumber <= 9] <-

paste0("0",as.character(allseasons$sznnumber[allseasons$sznnumber <= 9]))

allseasons$seasonnumchar[allseasons$sznnumber > 9] <-

as.character(allseasons$sznnumber[allseasons$sznnumber > 9])

# for sorting reasons, have the placement as a character

allseasons$placementchar[allseasons$placement <= 9] <-

paste0("0",as.character(allseasons$placement[allseasons$placement <= 9]))

allseasons$placementchar[allseasons$placement > 9] <-

as.character(allseasons$placement[allseasons$placement > 9])

# Distribution of scores at the end of the competition/seasons

graphalltitles <- function(dataset) {

dataset %>%

ggplot(aes(x=placement,y=indexWeight,label=chef)) +

facet_wrap(~paste0("Season ",seasonnumchar)) +

geom_hline(yintercept=0,color="#ffbc69")+

geom_text(hjust=0.5,size=2) +

theme_minimal() +

labs(title=paste0("Top Chef Weighted Index Scores at the End of Seasons")

,subtitle="Comparing All Chefs Across All Seasons\n")+

ylab("Index Score") + xlab("Placement") +

scale_y_continuous(lim=c(-25,50)) +

scale_x_continuous(lim=c(0,20),breaks=seq(1,18,2),labels = seq(1,18,2)) +

theme(panel.grid = element_blank()

,axis.text.x=element_text(size=6,color="black")

,axis.text.y=element_text(size=6,color="black")

,axis.ticks=element_line(color="gray15")

,axis.line=element_line(color="gray15")

,strip.background=element_rect(fill="darkcyan")

,strip.text=element_text(color="black"))

}

graphnotitlenocaption <- function(dataset) {

dataset %>%

ggplot(aes(x=placement,y=indexWeight,label=chef)) +

facet_wrap(~paste0("Season ",seasonnumchar)) +

geom_hline(yintercept=0,color="#ffbc69")+

geom_text(hjust=0.5,size=2) +

theme_minimal() +

ylab("Index score") + xlab("Placement") +

scale_y_continuous(lim=c(-25,50)) +

scale_x_continuous(lim=c(0,20),breaks=seq(1,18,2),labels = seq(1,18,2)) +

theme(panel.grid = element_blank()

,axis.text.x=element_text(size=6,color="black")

,axis.text.y=element_text(size=6,color="black")

,axis.ticks=element_line(color="gray15")

,axis.line=element_line(color="gray15")

,strip.background=element_rect(fill="darkcyan")

,strip.text=element_text(color="black"))

}

graphonlycaption <- function(dataset) {

dataset %>%

ggplot(aes(x=placement,y=indexWeight,label=chef)) +

facet_wrap(~paste0("Season ",seasonnumchar)) +

geom_hline(yintercept=0,color="#ffbc69")+

geom_text(hjust=0.5,size=2) +

theme_minimal() +

ylab("Index Score") + xlab("Placement") +

scale_y_continuous(lim=c(-25,50)) +

scale_x_continuous(lim=c(0,20),breaks=seq(1,18,2),labels = seq(1,18,2)) +

theme(panel.grid = element_blank()

,axis.text.x=element_text(size=6,color="black")

,axis.text.y=element_text(size=6,color="black")

,axis.ticks=element_line(color="gray15")

,axis.line=element_line(color="gray15")

,strip.background=element_rect(fill="darkcyan")

,strip.text=element_text(color="black")) +

labs(caption="Scoring: Elimination win = 7 points. Elimination high = 3. Elimination low = -3. Eliminated = -7.\nQuickfire win = 4. Quickfire high = 2. Quickfire low = -2.\nData github.com/celevitz/topChef ||| Twitter @carlylevitz")

}

graphalltitles(allseasons %>%

filter(sznnumber %in% c(1,2,3,4,5,6)) )

graphnotitlenocaption(allseasons %>%

filter(sznnumber %in% c(7,8,9,10,11,12)) )

graphnotitlenocaption(allseasons %>%

filter(sznnumber %in% c(13,14,15,16,17,18)) )

graphonlycaption(allseasons %>%

filter(sznnumber %in% c(19,20)) )

```

##### 5.b.ii. Code

```{r Viz_FinalIndex_Names_Code, eval=FALSE,echo=TRUE,message=FALSE,warning=FALSE}

library(topChef)

library(ggplot2)

library(tidyr)

library(dplyr)

## Get the index for all seasons

allseasons <- weightedindex("US",1,20,20)

for (season in seq(2,20,1)) {

allseasons <- rbind(allseasons,weightedindex("US",season,20,20))

}

## Graph it

# for sorting reasons, have the season be a character

allseasons$seasonnumchar[allseasons$sznnumber <= 9] <-

paste0("0",as.character(allseasons$sznnumber[allseasons$sznnumber <= 9]))

allseasons$seasonnumchar[allseasons$sznnumber > 9] <-

as.character(allseasons$sznnumber[allseasons$sznnumber > 9])

# for sorting reasons, have the placement as a character

allseasons$placementchar[allseasons$placement <= 9] <-

paste0("0",as.character(allseasons$placement[allseasons$placement <= 9]))

allseasons$placementchar[allseasons$placement > 9] <-

as.character(allseasons$placement[allseasons$placement > 9])

# Distribution of scores at the end of the competition/seasons

graphalltitles <- function(dataset) {

dataset %>%

ggplot(aes(x=placement,y=indexWeight,label=chef)) +

facet_wrap(~paste0("Season ",seasonnumchar)) +

geom_hline(yintercept=0,color="#ffbc69")+

geom_text(hjust=0.5,size=2) +

theme_minimal() +

labs(title=paste0("Top Chef Weighted Index Scores at the End of Seasons")

,subtitle="Comparing All Chefs Across All Seasons\n")+

ylab("Index Score") + xlab("Placement") +

scale_y_continuous(lim=c(-25,50)) +

scale_x_continuous(lim=c(0,20),breaks=seq(1,18,2),labels = seq(1,18,2)) +

theme(panel.grid = element_blank()

,axis.text.x=element_text(size=6,color="black")

,axis.text.y=element_text(size=6,color="black")

,axis.ticks=element_line(color="gray15")

,axis.line=element_line(color="gray15")

,strip.background=element_rect(fill="darkcyan")

,strip.text=element_text(color="black"))

}

graphnotitlenocaption <- function(dataset) {

dataset %>%

ggplot(aes(x=placement,y=indexWeight,label=chef)) +

facet_wrap(~paste0("Season ",seasonnumchar)) +

geom_hline(yintercept=0,color="#ffbc69")+

geom_text(hjust=0.5,size=2) +

theme_minimal() +

ylab("Index score") + xlab("Placement") +

scale_y_continuous(lim=c(-25,50)) +

scale_x_continuous(lim=c(0,20),breaks=seq(1,18,2),labels = seq(1,18,2)) +

theme(panel.grid = element_blank()

,axis.text.x=element_text(size=6,color="black")

,axis.text.y=element_text(size=6,color="black")

,axis.ticks=element_line(color="gray15")

,axis.line=element_line(color="gray15")

,strip.background=element_rect(fill="darkcyan")

,strip.text=element_text(color="black"))

}

graphonlycaption <- function(dataset) {

dataset %>%

ggplot(aes(x=placement,y=indexWeight,label=chef)) +

facet_wrap(~paste0("Season ",seasonnumchar)) +

geom_hline(yintercept=0,color="#ffbc69")+

geom_text(hjust=0.5,size=2) +

theme_minimal() +

ylab("Index Score") + xlab("Placement") +

scale_y_continuous(lim=c(-25,50)) +

scale_x_continuous(lim=c(0,20),breaks=seq(1,18,2),labels = seq(1,18,2)) +

theme(panel.grid = element_blank()

,axis.text.x=element_text(size=6,color="black")

,axis.text.y=element_text(size=6,color="black")

,axis.ticks=element_line(color="gray15")

,axis.line=element_line(color="gray15")

,strip.background=element_rect(fill="darkcyan")

,strip.text=element_text(color="black")) +

labs(caption="Scoring: Elimination win = 7 points. Elimination high = 3. Elimination low = -3. Eliminated = -7.\nQuickfire win = 4. Quickfire high = 2. Quickfire low = -2.\nData github.com/celevitz/topChef ||| Twitter @carlylevitz")

}

graphalltitles(allseasons %>%

filter(sznnumber %in% c(1,2,3,4,5,6)) )

graphnotitlenocaption(allseasons %>%

filter(sznnumber %in% c(7,8,9,10,11,12)) )

graphnotitlenocaption(allseasons %>%

filter(sznnumber %in% c(13,14,15,16,17,18)) )

graphonlycaption(allseasons %>%

filter(sznnumber %in% c(19,20)) )

```

### 5.c. Examples that use the `weightedindex` function 10 Elimination Challenges or 7 Quickfires into the season

##### 5.c.i. Visualizations

```{r Viz_IndexAllSeasons_SetUp, eval=TRUE,echo=FALSE,message=FALSE,include=FALSE,warning=FALSE}

library(topChef)

library(ggplot2)

library(tidyr)

library(dplyr)

## Get the index for all seasons

allseasons <- weightedindex("US",1,10,7)

for (season in seq(2,20,1)) {

allseasons <- rbind(allseasons,weightedindex("US",season,10,7))

}

```

The circles represent outliers: Michael in Las Vegas, Richard in All Stars: New York, Paul in Texas, and Kristen in Seattle. The thick bar within the rectangles are the median scores of chefs scores in that season.

```{r Viz_IndexAllSeasons, eval=TRUE,echo=FALSE,message=FALSE,warning=FALSE}

# drop unneeded variables

allseasons <- allseasons[,c("chef","szn","sznnumber","placement","indexWeight")]

## Graph it

# for sorting reasons, have the season be a character

allseasons$seasonnumchar[allseasons$sznnumber <= 9] <-

paste0("0",as.character(allseasons$sznnumber[allseasons$sznnumber <= 9]))

allseasons$seasonnumchar[allseasons$sznnumber > 9] <-

as.character(allseasons$sznnumber[allseasons$sznnumber > 9])

# for sorting reasons, have the placement as a character

allseasons$placementchar[allseasons$placement <= 9] <-

paste0("0",as.character(allseasons$placement[allseasons$placement <= 9]))

allseasons$placementchar[allseasons$placement > 9] <-

as.character(allseasons$placement[allseasons$placement > 9])

# Distribution of scores at this stage of the competition

allseasons %>%

ggplot(aes(x=seasonnumchar,y=indexWeight) ) +

geom_boxplot() +

# add horizontal line at 0

geom_hline(yintercept=0, color="#ffbc69") +

theme_minimal() +

labs(title=paste0("Top Chef Weighted Index: 7 quickfires & 10 elimination challenges\ninto each season")

,subtitle="Comparing All Chefs Across All Seasons\n"

,caption="Scoring: Elimination win = 7 points. Elimination high = 3. Elimination low = -3. Eliminated = -7.\nQuickfire win = 4. Quickfire high = 2. Quickfire low = -2.\nData github.com/celevitz/topChef ||| Twitter @carlylevitz")+

scale_x_discrete(labels=unique(allseasons$szn[order(allseasons$sznnumber)])) +

theme_minimal() +

ylab("Index score") + xlab("") +

theme(panel.grid = element_blank()

,axis.text.x=element_text(angle=90)

)

```

This example shows the index scores for the Top Four chefs in all seasons with the remaining six chefs from Top Chef World All Stars. Ali, Buddha, and Amar thus far have the highest index scores.

```{r Viz_IndexTopFour, eval=TRUE,echo=FALSE,message=FALSE,warning=FALSE}

# Top Four: Distribution of scores at this stage of the competition

allseasons[allseasons$placement <= 4,] %>%

ggplot(aes(x=placement,y=indexWeight,label=chef) ) +

# add horizontal line at 0

geom_hline(yintercept=0, color="#ffbc69") +

geom_text(size=2) +

theme_minimal() +

labs(title=paste0("Top Chef Weighted Index: 7 quickfires & 10 elimination challenges\ninto each season")

,subtitle="Comparing Top Four Chefs Across All Seasons\n"

,caption="Scoring: Elimination win = 7 points. Elimination high = 3. Elimination low = -3. Eliminated = -7.\nQuickfire win = 4. Quickfire high = 2. Quickfire low = -2.\nData github.com/celevitz/topChef ||| Twitter @carlylevitz")+

scale_x_continuous(lim=c(.5,4.5),breaks=c(1,1.5,2,3,4),labels=c("1","Current\nseason","2","3","4")) +

theme_minimal() +

ylab("Index score") + xlab("Placement") +

theme(panel.grid = element_blank()

)

```

##### 5.c.ii. Code

```{r Code_IndexAllSeasons, eval=FALSE,echo=TRUE,message=FALSE,warning=FALSE}

library(topChef)

library(ggplot2)

library(tidyr)

library(dplyr)

## Get the index for all seasons

allseasons <- weightedindex("US",1,10,7)

for (season in seq(2,20,1)) {

allseasons <- rbind(allseasons,weightedindex("US",season,10,7))

}

# drop unneeded variables

allseasons <- allseasons[,c("chef","szn","sznnumber","placement","indexWeight")]

## Graph it

# for sorting reasons, have the season be a character

allseasons$seasonnumchar[allseasons$sznnumber <= 9] <-

paste0("0",as.character(allseasons$sznnumber[allseasons$sznnumber <= 9]))

allseasons$seasonnumchar[allseasons$sznnumber > 9] <-

as.character(allseasons$sznnumber[allseasons$sznnumber > 9])

# for sorting reasons, have the placement as a character

allseasons$placementchar[allseasons$placement <= 9] <-

paste0("0",as.character(allseasons$placement[allseasons$placement <= 9]))

allseasons$placementchar[allseasons$placement > 9] <-

as.character(allseasons$placement[allseasons$placement > 9])

# Distribution of scores at this stage of the competition

allseasons %>%

ggplot(aes(x=seasonnumchar,y=indexWeight) ) +

geom_boxplot() +

# add horizontal line at 0

geom_hline(yintercept=0, color="#ffbc69") +

theme_minimal() +

labs(title=paste0("Top Chef Weighted Index: 7 quickfires & 10 elimination challenges\ninto each season")

,subtitle="Comparing All Chefs Across All Seasons\n"

,caption="Scoring: Elimination win = 7 points. Elimination high = 3. Elimination low = -3. Eliminated = -7.\nQuickfire win = 4. Quickfire high = 2. Quickfire low = -2.\nData github.com/celevitz/topChef ||| Twitter @carlylevitz")+

scale_x_discrete(labels=unique(allseasons$szn[order(allseasons$sznnumber)])) +

theme_minimal() +

ylab("Index score") + xlab("") +

theme(panel.grid = element_blank()

,axis.text.x=element_text(angle=90)

)

# Top Four: Distribution of scores at this stage of the competition

allseasons[allseasons$placement <= 4,] %>%

ggplot(aes(x=placement,y=indexWeight,label=chef) ) +

# add horizontal line at 0

geom_hline(yintercept=0, color="#ffbc69") +

geom_text(size=2) +

theme_minimal() +

labs(title=paste0("Top Chef Weighted Index: 7 quickfires & 10 elimination challenges\ninto each season")

,subtitle="Comparing Top Four Chefs Across All Seasons\n"

,caption="Scoring: Elimination win = 7 points. Elimination high = 3. Elimination low = -3. Eliminated = -7.\nQuickfire win = 4. Quickfire high = 2. Quickfire low = -2.\nData github.com/celevitz/topChef ||| Twitter @carlylevitz")+

scale_x_continuous(lim=c(.5,4.5),breaks=c(1,1.5,2,3,4),labels=c("1","Current\nseason","2","3","4")) +

theme_minimal() +

ylab("Index score") + xlab("Placement") +

theme(panel.grid = element_blank()

)

```