NAMCS08-Analysis.Rmd

---
title: "NAMCS-2008 Analysis with Shiny"
author: "shramesh"
#date: "8/21/2021"
output: 
  html_document:
    toc: TRUE
    toc_float: TRUE
runtime: shiny
---

```{r include=FALSE}
library(tidyverse)
library(shiny)
library(rlang) # to use !!parse_expr()
load('Data/namcs08.RData') #National Ambulatory Medical Care Survey (NAMCS) data

```

### Introduction

###### This shiny document is created using NAMCS 2008 dataset and I have used shiny widgets to create some plots and tables to understand some of the information in the dataset.

### Mean Systolic Blood Pressure (SBP)

```{r echo=FALSE}
selectInput('diagnosis', label = 'Which diagnosis would you like to explore?',
            choices = c('Diabetes'='diabetes',
                        'Congestive Heart Failure' ='chf',
                        'Hypertension'= 'htn'))

selectInput('age_filter', label = 'For everyone?',
       choices = c('Yes'='TRUE',
                  'Age greater than 65'=  'age > 65'))


renderTable({
left_join(patients,pmh) %>% 
  filter(!!parse_expr(input$age_filter)) %>% 
  group_by(!!parse_expr(input$diagnosis)) %>% 
  summarize(meanSBP = mean(sbp, na.rm = TRUE))
})
```

### Top reasons for visit
```{r echo=FALSE}
# sliderInput -> we don't have to parse_expr()
sliderInput('rows_to_show',
            label = 'How many rows to show?',
            min = 1,
            max = 10,
            step = 1,
            value = 5)

renderTable({
patients %>%
  group_by(Sex = sex,
           `Visit Reason` = visitreason) %>%
  summarize(count=n()) %>%
  arrange(desc(count)) %>%
  group_by(Sex) %>%
  slice(1:input$rows_to_show)
})
```

### Insurance used by patients

```{r echo=FALSE }
  patients %>% 
  ggplot(aes(x=fct_infreq(paytype))) +
  geom_bar()+
  labs(x="Insurance Type", y="Frequency") +
  coord_flip()


```

### Relation between age and height/weight

```{r echo=FALSE}
selectInput('yvar', label = 'Which variable on y axis?',
            choices = c('height','weight'))

sliderInput('age_range',
            label = 'What is the age range?',
            min = 0, max = 100,
            value = c(0,100),
            step = 10)

checkboxInput('smoothed_line',
              label = 'Should we show the smoothed line?',
              value = FALSE)

renderPlot({
age_plot = patients %>%
  ggplot(aes(x=age,y=!!parse_expr(input$yvar))) +
  geom_point() + 
  coord_cartesian(xlim = c(input$age_range[1], input$age_range[2]))

if(input$smoothed_line) {  #if_else will not work with ggplot
  age_plot + geom_smooth()
} else {
  age_plot
}

})

```


### Relation between depression and diabetes

```{r echo=FALSE}
selectInput('x_mapping',
            label = 'What is on the x axis?',
            choices = c('Diabetes'='diabetes',
                        'Depression'='depression'))

selectInput('fill_mapping',
            label = 'What is being filled?',
            choices = c('Depression'='depression'
                        ,'Diabetes'='diabetes'))

renderPlot({
pmh %>% 
  mutate(diabetes = if_else(diabetes == 1, 'Diabetes','No diabetes')) %>% 
  mutate(depression = if_else(depression == 1, 'Depression','No depression')) %>%
  ggplot(aes(x=!!parse_expr(input$x_mapping), 
             fill=!!parse_expr(input$fill_mapping))) +
            geom_bar(position='fill') +
    theme_minimal()+
    scale_fill_manual(values=c("darkcyan","darksalmon")) +
    labs(y="Proportion") 
    
})


```