AYA_Codeathon_Visualization_Day4_v02.Rmd

---
title: "AYA_Codeathon_Visualization_Day4_v0.2"
author: "Faith Lee and Lesley Chapman"
output: pdf_document
---

## DATA VISUALIZATION FUNCTIONS 
```{r setup, include=FALSE}
library(ggplot2)
library(dplyr)
library(usmap)
SEER_Cleaned <- read.csv("codeathon_RECODE_v03.csv")
```


## TOP 10 CANCERS 
```{r}
heatmap_data<- SEER_Cleaned %>% 
  group_by(Year.of.diagnosis, Site.recode.ICD.O.3.WHO.2008) %>% 
  summarize(cancer_count = n()) %>% arrange(Year.of.diagnosis, desc(cancer_count)) %>% 
  group_by(Year.of.diagnosis) %>% slice(1:10) 

ggplot(heatmap_data, aes(x=factor(Year.of.diagnosis), y=Site.recode.ICD.O.3.WHO.2008, fill= cancer_count)) + 
  geom_tile() + 
  xlab("Year of Diagnosis") + 
  ylab("Type of Cancer") + 
  ggtitle("Top 10 cancers of each year")
```

## BY AGE GROUP
```{r}
heatmap_data<- SEER_Cleaned %>% 
  group_by(Year.of.diagnosis, Age.recode.with..1.year.olds) %>% 
  summarize(cancer_count = n())

ggplot(heatmap_data, aes(x=factor(Year.of.diagnosis), y=Age.recode.with..1.year.olds, fill= cancer_count)) + 
  geom_tile(color="white") + 
  xlab("Year of Diagnosis") + 
  ylab("Age Group") + 
  ggtitle("AYA Cancer Counts by Age Group") + 
  scale_fill_gradient(low="pink", high="red")+theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))
```

## BY RACE 
```{r}
heatmap_data<- SEER_Cleaned %>% 
  group_by(Year.of.diagnosis, Race.and.origin.recode..NHW..NHB..NHAIAN..NHAPI..Hispanic.) %>% 
  summarize(cancer_count = n()) 

ggplot(heatmap_data, aes(x=factor(Year.of.diagnosis), y=Race.and.origin.recode..NHW..NHB..NHAIAN..NHAPI..Hispanic., fill= cancer_count)) + 
  geom_tile() + 
  xlab("Year of Diagnosis") + 
  ylab("Race And Origin") + 
  ggtitle("AYA Cancer Counts by Race and Origin") + 
  scale_fill_gradient(low="chartreuse", high="chartreuse4")+theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))
```
## IS there a relation between education level and cancer count? 
## Distribution of Survival Months BY RACE 

```{r}
ggplot(SEER_Cleaned_updated, aes(x=Race.and.origin.recode..NHW..NHB..NHAIAN..NHAPI..Hispanic., y=Survival.months, fill=factor(Decade))) + 
  geom_boxplot() + 
  ggtitle("Survival Months by Race") + 
  xlab("Race and Origin") +ylab("Survival Months") + 
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))
```
#### Survival Plots
## Survival time between men and women .
```{r}
ggplot(SEER_Cleaned_updated, aes(x=Survival.months, color=Sex)) +
  geom_histogram(fill="white", alpha=0.5, position="identity") + 
  ylab("Frequency") + 
  xlab("Survival Months") + 
  ggtitle("Distribution of Survival months by Gender") + 
  theme_bw()

# map_data <- SEER_Cleaned_updated %>% group_by(State) %>% summarize(count=n())
# names(map_data)[names(map_data) == 'State'] <- 'state'
# plot_usmap(data = map_data, values="count", color="blue") + 
#   labs(title = "U.S. States", subtitle = "Map of the United States") + 
#   theme(panel.background=element_blank(), legend.position = "right") +
#   scale_fill_continuous(low = "white", high = "red", name = "Cancer Counts \n 1997-2016 in 13 States")
```

### Note: Survival.months.flag is currently coded as 1 not 'complete'
**1997 to 2006**
```{r}
SEER_1997_2006 <- SEER_Cleaned_updated %>%
  filter(Year.of.diagnosis >= 1997 & Year.of.diagnosis <= 2006) %>%
  filter(Survival.months > 0)
```

```{r}
er <-  SEER_1997_2006 %>% distinct(Year.of.diagnosis)
er$Year.of.diagnosis
```



```{r}
ggplot(SEER_1997_2006, aes(x=Survival.months, color=Unemployed_cat)) +
  geom_histogram(fill="white", alpha=0.5, position="identity") + 
  ylab("Frequency") + 
  xlab("Survival Months") + 
  ggtitle("1997-2006: Distribution of Survival months by Employment Status") + 
  theme_bw() +
  guides(col=guide_legend("Employment_Status"))
```



```{r}
ggplot(SEER_1997_2006, aes(x=Survival.months, color=Rural.Urban.Continuum.Code.2013)) +
  geom_histogram(fill="white", alpha=0.5, position="identity") + 
  ylab("Frequency") + 
  xlab("Survival Months") + 
  ggtitle("1997-2006: Distribution of Survival months by Region : Rural vs Urban") + 
  theme_bw() +
  guides(col=guide_legend("Rural_Urban"))
```
```{r}
ggplot(SEER_1997_2006, aes(x=Survival.months, color=HighSchoolEdCat)) +
  geom_histogram(fill="white", alpha=0.5, position="identity") + 
  ylab("Frequency") + 
  xlab("Survival Months") + 
  ggtitle("1997-2006: Distribution of Survival months by High School Education") + 
  theme_bw() +
  guides(col=guide_legend("Education Stauts"))
```

**2007-2016**
```{r}
SEER_2007_2016 <- SEER_Cleaned_updated %>%
  filter(Year.of.diagnosis >= 2007  & Year.of.diagnosis <= 2016) %>%
  filter(Survival.months > 0)
```



```{r}
er <-  SEER_2007_2016 %>% distinct(Year.of.diagnosis)
er$Year.of.diagnosis
```

```{r}
ggplot(SEER_2007_2016, aes(x=Survival.months, color=Unemployed_cat)) +
  geom_histogram(fill="white", alpha=0.5, position="identity") + 
  ylab("Frequency") + 
  xlab("Survival Months") + 
  ggtitle("2007-2016: Distribution of Survival months by Employment Status") + 
  theme_bw() +
  guides(col=guide_legend("Employment_Status"))
```
```{r}
ggplot(SEER_2007_2016, aes(x=Survival.months, color=Rural.Urban.Continuum.Code.2013)) +
  geom_histogram(fill="white", alpha=0.5, position="identity") + 
  ylab("Frequency") + 
  xlab("Survival Months") + 
  ggtitle("2007-2016: Distribution of Survival months by Region : Rural vs Urban") + 
  theme_bw() +
  guides(col=guide_legend("Rural_Urban"))
```
```{r}
ggplot(SEER_2007_2016, aes(x=Survival.months, color=HighSchoolEdCat)) +
  geom_histogram(fill="white", alpha=0.5, position="identity") + 
  ylab("Frequency") + 
  xlab("Survival Months") + 
  ggtitle("2007-2016: Distribution of Survival months by High School Education") + 
  theme_bw() +
  guides(col=guide_legend("Education Stauts"))
```

## MOST COMMON CANCERS TO MEN 
## MOST COMMON CANCERS TO WOMEN 
## CANCER COUNTS by unemployedment year over year 
```{r}
lineplot <- SEER_Cleaned_updated%>% group_by(Year.of.diagnosis, Sex) %>% summarize(count=n())
ggplot(data=lineplot, aes(x=Year.of.diagnosis, y=count, group=Sex)) +
  geom_line(aes(linetype=Sex))+ theme_bw() +
  geom_point(aes(shape=Sex))+ ggtitle("AYA Counts by Gender") + 
  ylab("Count") + 
  xlab("Year")

ggplot(SEER_Cleaned_updated, aes(x=Chemotherapy.recode..yes..no.unk., y=Survival.months, fill=factor(Decade))) + 
  geom_boxplot() + 
  ggtitle("Survival Months by Chemotherapy") + 
  xlab("Chemotherapy (Yes, No| Unknown)") +
  ylab("Survival Months") + 
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))

```


## REMOVAL OF 31307 rows 
```{r}
ggplot(SEER_Cleaned_updated, aes(x=factor(median_income_household_group), y=Survival.months, fill=factor(Decade))) + 
  geom_boxplot() + 
  ggtitle("Survival Months by Median Income Household Group") + 
  xlab("Median Income Group") +ylab("Survival Months") + 
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))

ggplot(SEER_Cleaned_updated, aes(x=factor(Unemployed_cat), y=Survival.months, fill=factor(Decade))) + 
  geom_boxplot() + 
  ggtitle("Survival Months by Unemployment in County") + 
  xlab("Unemployment Level In County") +ylab("Survival Months") + 
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))
```


## POVERTY, SURVIVAL AND COUNTS 
```{r}
SEER_Cleaned_updated$family_poverty<- as.double(SEER_Cleaned_updated$X..Families.below.poverty.ACS.2013.2017)/100
tt<-SEER_Cleaned_updated  %>% 
  group_by(family_poverty, Race.and.origin.recode..NHW..NHB..NHAIAN..NHAPI..Hispanic.)%>% 
  summarize(count=n(), median_survival_months = median(Survival.months, na.rm=TRUE))

tt$family_poverty_grouped<- cut(tt$family_poverty, breaks = c(0,10,20,30,40,50,100))
tt_count <- tt%>% filter(count>100)

ggplot(tt_count, aes(x=family_poverty,median_survival_months)) + 
  geom_point(aes(size = count, color=Race.and.origin.recode..NHW..NHB..NHAIAN..NHAPI..Hispanic.)) + 
  ggtitle("Counts by Median Survival Months and Family Poverty (>200) w Racial component") + 
  ylab("Median Survival") + xlab("Family Poverty (%) at County Level") + 
  theme_bw()
```

## TOP 10 Cancer 
```{r}
SEER_Cleaned_updated %>% 
  filter(Survival.months.flag==1) %>% 
  summarize(median(Survival.months, na.rm=TRUE))
```