verification (version 1.44)

reliability.plot: Reliability Plot

Description

A reliability plot is a simple form of an attribute diagram that depicts the performance of a probabilistic forecast for a binary event. In this diagram, the forecast probability is plotted against the observed relative frequency. Ideally, this value should be near to each other and so points falling on the 1:1 line are desirable. For added information, if one or two forecasts are being verified, sharpness diagrams are presented in the corners of the plot. Ideally, these histograms should be relatively flat, indicating that each bin of probabilities is use an appropriate amount of times.

Usage

# S3 method for default
reliability.plot(x, obar.i, prob.y, titl = NULL, legend.names = NULL, ... )
# S3 method for verify
reliability.plot(x, ...)

Arguments

x

Forecast probabilities.(\(y_i\)) or a ``prob.bin'' class object from verify.

obar.i

Observed relative frequency \(\bar{o}_i\).

prob.y

Relative frequency of forecasts

titl

Title

legend.names

Names of each model that will appear in the legend.

...

Graphical parameters.

Author

Matt Pocernich

Details

This function works either by entering vectors or on a verify class object.

References

Wilks, D. S. (1995) Statistical Methods in the Atmospheric Sciences Chapter 7, San Diego: Academic Press.

Examples

## Data from Wilks, table 7.3 page 246.
 y.i   <- c(0,0.05, seq(0.1, 1, 0.1))
 obar.i <- c(0.006, 0.019, 0.059, 0.15, 0.277, 0.377, 0.511,
    0.587, 0.723, 0.779, 0.934, 0.933)

 prob.y <- c(0.4112, 0.0671, 0.1833, 0.0986, 0.0616, 0.0366,
    0.0303,  0.0275, 0.245, 0.022, 0.017, 0.203) 

 obar <- 0.162

reliability.plot(y.i, obar.i, prob.y, titl = "Test 1", legend.names =
c("Model A") )


## Function will work with a ``prob.bin'' class object as well.
## Note this is a very bad forecast.
obs<- round(runif(100))
pred<- runif(100)

A<- verify(obs, pred, frcst.type = "prob", obs.type = "binary")

reliability.plot(A, titl = "Alternative plot")