---
title: "getting-started"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{getting-started}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r setup, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  eval = greta:::check_tf_version("message"),
  cache = TRUE,
  comment = "#>"
)
knitr::opts_knit$set(global.par = TRUE)
set.seed(1)
```

```{r library}
library(greta.gp)
```

### Gaussian processes in greta

`greta.gp` extends greta to let you define Gaussian processes as part of your model. It provides a syntax to create and combine GP kernels, and use them to define either full rank or sparse Gaussian processes.

#### Example

```{r simulate, message = FALSE}
# simulate data
x <- runif(20, 0, 10)
y <- sin(x) + rnorm(20, 0, 0.5)
x_plot <- seq(-1, 11, length.out = 200)
```

```{r model, message = FALSE}
library(greta)
library(greta.gp)

# hyperparameters
rbf_var <- lognormal(0, 1)
rbf_len <- lognormal(0, 1)
obs_sd <- lognormal(0, 1)

# kernel & GP
kernel <- rbf(rbf_len, rbf_var) + bias(1)
f <- gp(x, kernel)

# likelihood
distribution(y) <- normal(f, obs_sd)

# prediction
f_plot <- project(f, x_plot)
```

```{r fit, message = FALSE}
# fit the model by Hamiltonian Monte Carlo
m <- model(f_plot)
draws <- mcmc(m)
```

```{r plotting, fig.width = 10, fig.height = 6, dpi = 200}
# plot 200 posterior samples
plot(
  y ~ x,
  pch = 16,
  col = grey(0.4),
  xlim = c(0, 10),
  ylim = c(-2.5, 2.5),
  las = 1,
  fg = grey(0.7),
)

for (i in 1:200) {
  lines(
    draws[[1]][i, ] ~ x_plot,
    lwd = 2,
    col = rgb(0.7, 0.1, 0.4, 0.1)
  )
}
```