BMA in FEWS.

The present version of BMA within FEWS uses an R Package for Probabilistic Forecasting using Ensembles and Bayesian Model Averaging - "Ensemble BMA Package". The package was developed by Chris Fraley, Adrian E. Raftery, J. McLean Sloughter and Tilmann Gneiting at the University of Washington. This package is distributed under General Public License (version >= 2) .

R Package can be downloaded from CRAN R project website or directly from R package . However it is recomended that you download the version of ensembleBMA from the link given below instead since the version 3.0-3 is tested and used within FEWS. The new version which is available from the above mentioned link may or may not work flawlessly within FEWS.

Ensemble BMA Package Documentation.

For Ensemble BMA R package documentation, please refer to the online documentation on Cran R project website through this link Ensemble BMA documentation

Versions

FEWS uses R version 2.7.0.
Package: ensembleBMA, Version: 3.0-3, Date: 2008-07-21
Supporting Package chron, Version: 2.3-24, Date: 2008-07-18
Please Note: The ensemble BMA is an older version that is no longer supported by the original developers

To run BMA in FEWS, firstly install the correct version of R on the computer where BMA Module is running .
Copy the contents of Ensemble BMA ensembleBMA.zip and Chron chron.zip under library directory of R Package.
Please Note: Use only the package versions as mentioned above for running BMA Module in Delft-FEWS.

Systematic Diagram

Preprocessor

BMA Module preprocesser prepares the input for ensemble BMA R Package. Ensemble BMA R Package uses input as CSV format. The General Adapater configuration of Preprocessor is shown as below.

<executeActivity>
	<command>
		<className>nl.wldelft.fews.adapter.bma.BmaPreAdapter</className>
	</command>
	<arguments>
		<argument>%ROOT_DIR%</argument> <!-- root directory -->
		<argument>piOutputTimeSeries/bmainputL0.csv</argument> <!-- outputfile  -->
		<argument>%TIME0%</argument>  <!-- Time0 -->
		<argument>0</argument> <!-- Start of Lead time period in days  -->
		<argument>parameters.txt</argument>  <!-- Parameter file - each column represents a row -->
		<argument>piOutputTimeSeries/forecast0.csv</argument>  <!-- Number of (partly) complete Forecasts used for calculating the training period -->
	</arguments>
	<timeOut>4000000</timeOut>
</executeActivity>

The above configuration has to be repeated for different lead time. Make sure that name of output file accordingly changed.

BMA Module

BMA Module is a script written under R package which uses the ensembleBMA package written for R as briefly described above. The General Adapater configuration for running BMA Module is shown as below.

<executeActivity>
	<command>
		<executable>$R_EXE$</executable>
	</command>
	<arguments>
		<argument>--vanilla</argument>
		<argument>%ROOT_DIR%/config/BMA_FEWS_Script.R</argument>
		<argument>%ROOT_DIR%/piOutputTimeSeries/bmainputL0.csv</argument> <!-- inputfile  -->
		<argument>%ROOT_DIR%/piOutputTimeSeries/bmaoutputL0.qan</argument> <!-- outputfile qauntile -->
		<argument>%ROOT_DIR%/piOutputTimeSeries/bmaoutputW0.wie</argument> <!-- outputfile  weights-->
		<argument>%ROOT_DIR%/piOutputTimeSeries/bmaoutputB0.bia</argument> <!-- outputfile  bias-->
		<argument>0</argument> <!- input lead time in days (not used) -->
		<argument>%ROOT_DIR%/piOutputTimeSeries/forecast0.csv</argument> <!-- inputfile Forecast Length -->
	</arguments>
	<timeOut>1000000000</timeOut>
	<overrulingDiagnosticFile>%ROOT_DIR%/rscript.log</overrulingDiagnosticFile>
</executeActivity>

The above configuration has to be repeated for different lead time. Make sure that name of input and output files are accordingly changed.

Please note: $R_EXE$ is attribute which is defined in global.properties file as "R_EXE=C:/Program Files/R/R-2.7.0/bin/Rscript.exe"

BMA Module R Script

A typical BMA Module R Script can be downloaded from here.

The contents of BMA Module R script is briefly described as below.

--- Read Arguments
--- Check if files exists
--- Read Forecast Length file
--- Load Ensemble R
--- Read input data

-- Assign labels (hard coded - similar to parameter file) (R-Code - Make sure to update this line for your model)
labels <-c("SBK_MaxLob_DWD_GME_Q.fs","SBK_MaxLob_DWD_LM_Q.fs",...............)

--- Perform ensembleBMA analaysis  (R-Code)
enRData<-ensembleData(forecasts=rdata[,labels], dates=rdata$TIME, observations=rdata$OBS)
trainingrule=list(length=forecastlen,lag=1)
rDataBMA <- ensembleBMA(enRData,model="normal",trainingRule=trainingrule, control = controlBMAnormal(maxIter=20))

--- output Quantile to File
--- output Wieghts to File
--- output Bias to File

Please make sure that the line "labels <-c("SBK_MaxLob_DWD_GME_Q.fs","SBK_MaxLob_DWD_LM_Q.fs",...............)" is changed according to the number of models used.

Output of each BMA Module run are 3 files, with extension ...

*.wei -> weights
*.bia -> bias
*.qan -> quantiles - value for the next first forecast - (used only for checking)

Format of Weights file (*.wei)

forecast-date, weight for model one, weight for model 2 , .... and so on .... , sigma

Format of Bias file (*.bia)

B value for model 1, B value for model 2, ... and so on ...
A value for model 1, A value for model 2, ... and so on ...

Postprocessor

BMA Module postprocesser reads the output of R model prepares the data which is to be later imported into FEWS database. The General Adapater configuration of Postprocessor is shown as below.

<executeActivity>
	<command>
		<className>nl.wldelft.fews.adapter.bma.BmaPostAdapter</className>
	</command>
	<arguments>
		<argument>%ROOT_DIR%</argument> <!-- root directory -->
		<argument>piOutputTimeSeries</argument> <!-- outputDirectory  -->
		<argument>%TIME0%</argument>  <!-- Time0 -->
		<argument>3</argument>  <!-- max lead time in days -->
		<argument>parameters.txt</argument> <!-- Parameter file - each column represents a row -->
	</arguments>
	<timeOut>4000000</timeOut>
	<overrulingDiagnosticFile>%ROOT_DIR%/piDiagnostic.xml</overrulingDiagnosticFile>
</executeActivity>

The postprocessor uses the output of BMA Module run (i.e. quantiles, weights and bias) and the input to generate new forecasted timeseries + quantiles (10 , 25, 75 and 90) timeseries.

Generating Forecast Timeseries

The forecasted timeseries are generated using the weights, sigma and bias correction.

    FOR EACH models_i  (skip missing forecasts)

       BMA += weight_i * (bias_a_i * forecast_i + bias_b_i)
       sumweights += weight_i

    END
    BMA = BMA / sumweights


    Quantile_10 = BMA - 0.842 * sigma
    Quantile_25 = BMA - 0.675 * sigma
    Quantile_50 = BMA
    Quantile_75 = BMA + 0.675 * sigma
    Quantile_90 = BMA + 0.842 * sigma

Page tree

BMA in FEWS