---
title: "CondExp"
author: "D. Renard"
date: "11 mai 2021"
output: pdf_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
library(RGeostats)
rm(list=ls())
```

## Introduction

This file is meant to demonstrate the capabilities of the correlation function and its auxiliary features

## Data

We create a set of variables in a Db. This Db is created as a regular grid for better legibility. The first variable is created by sampling a Normal distribution whereas the second one is obtained by adding a Uniform noise around the first variable.

```{r Data Base}
db = db.create(nx=c(100,100))
db = db.add(db,var1=rnorm(db$nech))
db = db.add(db,var2=var1 + runif(db$nech,min=-1,max=1))
```

## Correlation

The next plot shows the correlation between the two variables The correlation is discretized within a fine grid covering both axes. Each pixel is painted with a color which correponds to the number of samples contained within this pixel.

```{r Correlation}
correlation(db,name1="var1",name2="var2")
```

The previous figure can be managed as far as its dimensions are concerned. For example, we can decide to set both axes equal.


```{r Iso}
correlation(db,name1="var1",name2="var2",flag.iso=TRUE,asp=1)
```

## Adding some decorations

Adding the regression line in blue

```{r Regression_Line}
correlation(db,name1="var1",name2="var2", flag.regr=TRUE, reg.lwd=2,reg.col="blue")
```

Adding the first bisector in red (even if it does not make too much sense here). The plot is represented with same scale on both axes.

```{r First_Bisector}
correlation(db,name1="var1",name2="var2", flag.iso=TRUE,flag.diag=TRUE, diag.lwd=2,
              diag.col="red",asp=1)
```

Adding the regression curve in green

```{r Regression_Curve}
correlation(db,name1="var1",name2="var2", flag.ce=TRUE, nbr.ce=20, ce.lwd=4, ce.col="green")
```

## Representing the correlation as points

When the number of samples is not too large, we may prefer to represent the scatter plot as a set of points. Each sample is now represented by an individual point whose color is set equal to the sample rank

```{r As_Points}
correlation(db,name1="var1",name2="var2", ,name.col="rank",flag.aspoint=TRUE)
```