---
title: "Add a point to a scatterplot"
author: "Doug Hemken"
date: "October 2015"
output: 
  html_document:
    includes:
      after_body: ../Rmd/bottomRmdKB.html
      before_body: ../Rmd/topRmdKB.html
      in_header: ../Rmd/headRmdKB.html
    self_contained: no
    theme: null
    toc: yes
---

[Stata notes](../stata.html)

There are a couple of approaches one could take to
add a single point to a scatterplot.  One is to
overlay the scatterplot with the plot produced by
`scatteri`, an immediate scatterplot.

In this example, we will plot the overall mean in
both the $x$ and the $y$ variables.  A linear
regression of $y$ on $x$ always passes through this point.
A regression with higher-order terms seldom passes
through this point!

```{r, echo=FALSE, message=FALSE}
  require(knitr)
if (file.exists("C:/Program Files (x86)/Stata14/StataMP-64.exe")) {
  statapath <- "C:/Program Files (x86)/Stata14/StataMP-64.exe"
  } else if (file.exists("C:/Program Files (x86)/Stata14/StataSE-64.exe")) {
    statapath <- "C:/Program Files (x86)/Stata14/StataSE-64.exe"
    }
opts_chunk$set(engine="stata", engine.path=statapath, results="hide", comment="")
source_hook <- knit_hooks$get("source")
knit_hooks$set(source = function(x, options) {
    y <- strsplit(x, "\n")[[1]]
    # Find and remove graph export in Stata source
    graphexport <- grep("^graph export.*", y)
    if (length(graphexport)>0) {y <- y[-(graphexport)]}
    # Now treat the result as regular source code
    source_hook(y, options)
})
  writeLines(c("sysuse auto"), "profile.do")
```

## Identifying the mean

First we identify the mean values of $y$ and $x$ and save them
as `local` macro variables.

```{r}
  sysuse auto
  summarize price, meanonly
  local X = r(mean)
  summarize mpg, meanonly
  local Y = r(mean)
```

## Overlay scatter and scatteri
Then we overlay the scatterplot and the immediate scatterplot
of the single point.

```{r, echo=5}
  summarize price, meanonly
  local X = r(mean)
  summarize mpg, meanonly
  local Y = r(mean)
  twoway (scatter mpg price) (scatteri `Y' `X', msymbol(D))
  graph export "addpoint_files/gm1.png", replace
```

The `msymbol(D)` gives us a large, diamond-shaped point marker.

![scatter and scatteri](Addpoint_files/gm1.png)

## Label the point, add a regression line

```{r, echo=5:6}
  summarize price, meanonly
  local X = r(mean)
  summarize mpg, meanonly
  local Y = r(mean)
  twoway (scatter mpg price)(lfit mpg price) ///
    (scatteri `Y' `X' (6) "Grand Mean", msymbol(D))
  graph export "addpoint_files/gm2.png", replace
```

The `"(6)"` is a clock position for the point label.

![With label and regression](Addpoint_files/gm2.png)

## Use a quadratic fit, add better annotation

```{r, echo=5:8}
  summarize price, meanonly
  local X = r(mean)
  summarize mpg, meanonly
  local Y = r(mean)
  twoway (scatter mpg price)(qfit mpg price) ///
    (scatteri `Y' `X' (6) "Grand Mean", msymbol(D)), ///
    xtitle("Price ($)") ytitle("Mileage (mpg)") ///
    legend(order(1 "Observed" 2 "Predicted" 3 "Grand Mean"))
  graph export "addpoint_files/gm3.png", replace
```

Here we can see that the regression line misses the grand
mean.

![Quadratic fit and annotation](Addpoint_files/gm3.png)

```{r, engine='R', echo=FALSE, message=FALSE}
  unlink("profile.do")
```