# Changes between Version 3 and Version 4 of udg/ecoms/RPackage/examples/visualization

Ignore:
Timestamp:
May 15, 2016 1:05:11 PM (6 years ago)
Comment:

--

### Legend:

Unmodified
 v3 }}} visualizeR uses its own particular classes for handling data. It is necessary to convert to the visualizeR classes before using the visualization functions: {{{#!text/R prd <- as.MrEnsemble(tx.forecast) class(prd) ## [1] "MrEnsemble" ## attr(,"package") ## [1] "visualizeR" obs <- as.MrGrid(obsintp) class(obs) ## [1] "MrGrid" ## attr(,"package") ## [1] "visualizeR" }}} We next show different forecast visualization plots: For each member, the daily predictions are averaged to obtain a single seasonal forecast (this yields a first warning, as in this example). For rectangular spatial domains (i.e., for grids), the spatial average is first computed (with a warning) to obtain a unique series for the whole domain, as in this example. The corresponding terciles for each ensemble member are then computed for the analysis period. Thus, data is converted converted to a series of tercile categories by considering values above, between or below the terciles of the whole period. The probability of a member to fall into the observed tercile is represented by the colorbar (different color palettes are available through the argument color.pal). For instance, probabilities below 1/3 are very low, indicating that a minority of the members falls in the tercile. Conversely, probabilities above 2/3 indicate a high level of member agreement (more than 66% of members falling in the same tercile). The observed terciles (the events that actually occurred) are represented by the white circles. For each member, the daily predictions are averaged to obtain a single seasonal forecast (this yields a first warning, as in this example). For rectangular spatial domains (i.e., for grids), the spatial average is first computed (with a warning) to obtain a unique series for the whole domain, as in this example. The corresponding terciles for each ensemble member are then computed for the analysis period. Thus, data is converted to a series of tercile categories by considering values above, between or below the terciles of the whole period. The probability of a member to fall into the observed tercile is represented by the colorbar (different color palettes are available through the argument color.pal). For instance, probabilities below 1/3 are very low, indicating that a minority of the members falls in the tercile. Conversely, probabilities above 2/3 indicate a high level of member agreement (more than 66% of members falling in the same tercile). The observed terciles (the events that actually occurred) are represented by the white circles. Finally, the ROC Skill Score (ROCSS) is indicated in the secondary (right) Y axis. For each tercile, it provides a quantitative measure of the forecast skill, and it is commonly used to evaluate the performance of probabilistic systems. The value of this score ranges from 1 (perfect forecast system) to -1 (perfectly bad forecast system). A value zero indicates no skill compared with a random prediction. Using the bubble plot allows for instance to test the effrect of data detrending, as we suggested in the previous example that may be the source of some -artificial- skill in the north-eastern sector of the study area. We control this through the logical argument detrend: Using the bubble plot allows for instance to test the effect of data detrending, as we suggested in the previous example that may be the source of some -artificial- skill in the north-eastern sector of the study area. We control this through the logical argument detrend: