week1_wcgs_r_lab_workbook.Rmd

---
title: "WCGS data lab workbook"
author: "Maureen Lahiff"
date: "March 6, 2020"
---

# this R Markdown file assumes you went through the Week 1 R Tutorial first


```{r set libraries and read in data, include=FALSE}
library(ggplot2)
library(dplyr)
library(readr)
library(epitools)
library(readr)
library(rmarkdown)
library(knitr)

data(wcgs)
```

```{r variables we made in the Week 1 R Tutorial, include=FALSE}
wcgs$dibpat0_fact <- factor(wcgs$dibpat0, ordered = TRUE, labels = c("A","B"))


wcgs$smoker0[wcgs$ncigs0 > 0] <- 1
wcgs$smoker0[wcgs$ncigs0 == 0] <- 0

wcgs$smoker0_fact <- factor(wcgs$smoker0, ordered = TRUE, labels = c("Non-Smoker","Smoker"))


wcgs$highsbp0[wcgs$sbp0 >= 140] <- 1
wcgs$highsbp0[wcgs$sbp0 < 140] <- 0


wcgs$heightcm0 <- round(wcgs$height0 * 2.54, digits = 2)
wcgs$weightkg0 <- round(wcgs$weight0/2.2, digits = 2)
wcgs$BMI0 <- round(wcgs$weightkg0/((wcgs$heightcm0/100)^2), digits = 1)
```


```{r average cholesterol for smokers and non-smokers }

# Question 1 to turn in

# average cholesterol at baseline for smokers and non-smokers


wcgs %>% group_by(smoker0_fact) %>% summarize(average = mean(sbp0))
```
```{r}
head(wcgs)
```


```{r recode behavioral patterns}

# create a factor variable out of the 4-level behavioral pattern variable, behpat0

wcgs$behpat0_fact <- factor(wcgs$behpat0,
                           ordered = TRUE,
                           labels = c("A", "some A", "mix A and B","B"))
wcgs$smoker_example <- factor(labels = c("A","B"), wcgs$smoker0)
```

```{r create bmi categories}

# cut gives us a lot of flexibility 
# in our specifications of the interval endpoints 
# the intervals here are the commonly used ones

wcgs$bmi_cat <- cut(wcgs$BMI0,
                    breaks = c(0, 18.5, 25.0, 30.0, Inf),
                    include.lowest = TRUE,
                    right = FALSE,
                    ordered_results = TRUE,
                    labels = c("underweight", "normal", "overweight", "obese") )

```

```{r sbp and BMI}

wcgs %>% group_by(bmi_cat) %>% summarize(blood_pressure = mean(sbp0))

```

```{r}
```


```{r}
# control (command)+alt-i
# command+shift+c to make comments
```


```{r be sure chd at 8 1/2 yr follow-up is a factor}

wcgs$chd69_fact <- factor(wcgs$chd69, labels = c("No CHD", "CHD"))
```


```{r tables }

# these are examples for additional tables

behpat_table <- table(wcgs$behpat0_fact)
addmargins(behpat_table)
round(prop.table(behpat_table), digits = 3)



# table for behavioral pattern and chd

# the proportion option 1 asks for "row percents" in the two-way table 
# this is Question 2 to turn in

# making a two-way table chd and behavioral pattern
chd_behpat_table <- table(wcgs$behpat0_fact, wcgs$chd69)
addmargins(chd_behpat_table)
round(prop.table(chd_behpat_table, 1), digits = 3)

# making a two-way table chd and blood pressure
# 

wcgs$highsbp0_fact <- factor(wcgs$highsbp0, labels = c("High Sbp", "Low Sbp"))

chd_highsbp_table <- table(wcgs$highsbp0_fact, wcgs$chd69)
addmargins(chd_highsbp_table)
round(prop.table(chd_highsbp_table, 1), digits = 3)

```

```{r Question 4}

```


```{r side by side bar charts for chd69 by behavior pattern group}

# this is Question 3 to turn in

wcgs %>% count(behpat0_fact, chd69) %>%
  group_by(behpat0_fact) %>%
  mutate(prop = n/sum(n))  %>%
  ggplot(aes(x = behpat0_fact, y = prop, fill = chd69)) +
  geom_bar(width = 0.5, stat = "identity", position = "dodge") +
  labs(y = "percent", x = "behavior type", title = "Type A behavior and CHD at 8 1/2 year follow-up", 
       subtitle = "n = 3154 men at baseline") +
  theme(plot.title = element_text(hjust = 0.5), plot.subtitle = element_text(hjust = 0.5))
```

```{r summary, histogram, and box plots }

# summary for BMI by CHD at 8 /12 yr follow-up

no_chd_summary <- summary(subset(wcgs, chd69 == "No CHD", select = BMI0))

chd_summary <- summary(subset(wcgs, chd69 == "CHD", select = BMI0))

no_chd_summary
chd_summary

#use facet wrap to get side by side histgrams for baseline BMI by CHD

ggplot(wcgs, aes(x = BMI0, y = ..density..)) +
  geom_histogram(binwidth = 1, color = "black", fill = "purple") +
  facet_wrap(.~wcgs$chd69) + 
  labs(title = "WCGS baseline BMI by CHD group",
       subtitle = "men ages 39 to 59",
       x = "BMI ",
       y = "density" ) +
  scale_x_continuous(limits = c(18.5, 40),
                     breaks = c(20, 25, 30, 35),
                     labels = c(20, 25, 30, 35)) +
  theme(plot.title = element_text(hjust = 0.5),
        plot.subtitle = element_text(hjust = 0.5))

```