hydroGOF

About hydroGOF
SVN access
Download/Files
News
Check results
Package R docs

new

About hydroGOF

What is hydroGOF ?

hydroGOF is an R package that provides S3 functions implementing both statistical and graphical goodnes-of-fit measures between observed and simulated values, mainly oriented to be used during the calibration, validation, and application of hydrological models.

Missing values in observed and/or simulated values can removed before the computations.

Bugs / comments / questions / collaboration of any kind are very welcomed.

Installation

# Stable version install.packages("hydroGOF") # Latest (under-development) releaseinstall.packages("hydroGOF",, "http://rforge.net/", type="source")

Package Description

Package: hydroGOF Version: 0.3-6-3 Title: Goodness-of-fit functions for comparison of simulated and observed hydrological time series Author: Mauricio Zambrano-Bigiarini [aut, cre] Maintainer: Mauricio Zambrano Bigiarini Description: S3 functions implementing both statistical and graphical goodness-of-fit measures between observed and simulated values, mainly oriented to be used during the calibration, validation, and application of hydrological models. Missing values in observed and/or simulated values can be removed before computations. Comments / questions / collaboration of any kind are very welcomed. Depends: R (>= 2.10.0), zoo (>= 1.7-2) License: GPL (>=2)

Citation

citation("hydroGOF")

Vignette

http://cran.r-project.org/web/packages/hydroGOF/vignettes/hydroGOF_Vignette.pdf

Related Material

Comparing Goodness-of-fit Measures for Calibration of Models Focused on Extreme Events (EGU2012 abstract)
Comparing Goodness-of-fit Measures for Calibration of Models Focused on Extreme Events (EGU2012 poster)
R: a statistical environment for hydrological analysis (EGU2010 abstract)
R: a statistical environment for hydrological analysis (EGU2010 poster)

BugReports

https://www.rforge.net/bugzilla/describecomponents.cgi?product=hydroGOF

What's New?

NEWS/ChangeLog for hydroGOF -------------------------- 0.3-7 (under-development) o `valindex' : false warning fixed. This warning raised the message "`sim' and `obs' are empty or they do not have any common pair of elements with data !!", even when `sim' and `obs' were fully compatible. o `ggof' : -) new argument `gofs', which allows to select which GoFs are shown when 'gof.leg=TRUE' -) new argument `stype' -) new argument `season.names' -) `ftype' argument now allows `seasonal' value as input -) internal: 'hydroTSM::daily2monthly' was replaced by `daily2monthly' -) internal: 'hydroTSM::daily2annual' was replaced by `daily2annual' -) internal: 'hydroTSM::vector2zoo' was replaced by `vector2zoo' o `pbiasfdc' : -) internal: 'hydroTSM::fdc' was replaced by `fdc' o `plotbandsonly': -) internal: 'hydroTSM::vector2zoo' was replaced by `vector2zoo' o `plot2' : -) new argument `gofs', which allows to select which GoFs are shown when 'gof.leg=TRUE' -) internal: 'hydroTSM::vector2zoo' was replaced by `vector2zoo' -) internal: 'hydroTSM::drawTimeAxis' was replaced by `drawTimeAxis' o `plotbands' : -) internal: 'hydroTSM::vector2zoo' was replaced by `vector2zoo' -) internal: 'hydroTSM::drawTimeAxis' was replaced by `drawTimeAxis' ToDo: Import `zoo' package and remove all the 'zoo::' calls in the .R files 0.3-6 22-Mar-2013 o `gof' : -) now it works with matrices/data.frames of zoo/xts class. Thanks to L. Pagliero ! -) new arguments `s' and `method' which are passed to the `KGE' function (`out.type' is fixed to "single") -) new argument `j', which is passed to the `mNSE' function -) new argument `norm', which is passed to the `nrmse' function -) new arguments `lQ.thr' and `hQ.thr', which are passed to the (optional) `pbiasfdc' function o `NSE' : -) now it works with matrices/data.frames of zoo class. o `me' : -) now it works with matrices/data.frames of zoo class. o `mae' : -) now it works with matrices/data.frames of zoo class. o `mse' : -) now it works with matrices/data.frames of zoo class. o `rmse' : -) now it works with matrices/data.frames of zoo class. o `rsr' : -) now it works with matrices/data.frames of zoo class. o `rSD' : -) now it works with matrices/data.frames of zoo class. o `pbias' : -) now it works with matrices/data.frames of zoo class. o `mNSE' : -) now it works with matrices/data.frames of zoo class. o `rNSE' : -) now it works with matrices/data.frames of zoo class. o `md' : -) now it works with matrices/data.frames of zoo class. o `cp' : -) now it works with matrices/data.frames of zoo class. o `rPearson' : -) now it works with matrices/data.frames of zoo class. o `br2' : -) now it works with matrices/data.frames of zoo class. o `KGE' : -) now it works with matrices/data.frames of zoo class. o `VE' : -) now it works with matrices/data.frames of zoo class. o `d' : -) now it works with matrices/data.frames of zoo class. -) fixed typo in the formula presented in the pdf manual (a square was missing). Thanks to T. Ladson and I. Thiele-Eich ! o `rd' : -) now it works with matrices/data.frames of zoo class. -) fixed typo in the formula presented in the pdf manual (a square was missing). Thanks to I. Thiele-Eich ! o `nrmse' : -) now it works with matrices/data.frames of zoo class. o `nrmse.matrix': fixed error in checking the `norm' argument. o `valindex' : a warning was added when `sim' and `obs' do not have any pair of elements with data in both variables. o e-mail address of maintainer was changed from the @ing.unit.it to @gmail.com 0.3-5 19-Oct-2012 o now it depends on `hydroTSM' >= 0.3-6, in order to allow handling sub-daily values in `ggof' o `ggof' : -) now it handles sub-daily values in `sim' and `obs' (aggregation to daily/monthly/annual values not yet available) -) `dates' argument accepts `POSIXct' objects o `KGE' : -) two new arguments: `method', to allow KGE-2009 and KGE-2012 formulas, and `out.type', in order to allow look at the individual elements of the KGE (correlation, bias, and variability) -) Added reference to Kling et al. 2012 and improved documentation -) Added example to show difference between KGE-2009 and KGE-2012 -) now it returns NA instead of an error when there are no pairs of `sim' and `obs' without missing values (thanks to Giovanni Forzieri !) o DESCRIPTION : 'Author@R' field was replaced by 'Authors@R' 0.3-4 03-Jul-2012 o new function `VE', which computes the volumetric efficiency proposed by Criss and Winston 2008 (thanks to Eric Morway for giving the final push to include it !) o `gof' : VE is now included in the computed goodness-of-fit measures o `ggof' : VE is now included in the computed goodness-of-fit measures o `plot2' : VE is now included in the computed goodness-of-fit measures (when 'leg.gof=TRUE') o function `.intersect' was removed and function `valindex' was improved in order to replace it o `zoo' package was moved from `Imports' to `Depends' o `methods' package is now imported, in order to avoid problems with Rscript, which currently does not load `methods' 0.3-3 02-May-2012 o `gof' : added new example and description of the output class (thanks to Kris Jaeger !) o `br2' : the new example included in version 0.3-2 was improved o `ggof' : fixed problem with `ylim' when some elements in `sim' or `obs' were NA (the issue was in the `plot2' function) o `plot2' : fixed problem with `ylim' when some elements in `x' or `y' were NA 0.3-2 07-Oct-2011 o `br2' : fixed documentation error in its formula (thanks to Jan Corluy !), and a new example was added. Its value was already computed in the correct way. 0.3-1 14-Sep-2011 o `ggof' : fixed problem when `ylim' is specified by the user (the issue was in the `plot2' function) o `plot2' : fixed problem when `ylim' is specified by the user 0.3-0 01-Sep-2011 o NAMESPACE created, which leads to reduced dependency on the package hydroTSM. o hydroGOF now depends on xts >= 0.8-2 (=> zoo >= 1.7-2), in order to make use of improved plotting capabilities for ts objects. The first release of `xts' after the 0.8-0 (which was set as the default building block for plotting time series in hydroTSM) was the 0.8-2, and appeared on CRAN on August 9th, 2011 o now it depends on `hydroTSM' >= 0.3-0, in order to use the re-written 'drawxaxis/drawTimeAxis' function. o a vignette is now included with the package, in order to illustrate some basic functionality of hydroGOF o `plot2' : -) better display of time series objects, when `plot.type' is `single'. The (nicer) `plot.xts' function is used instead of the previous `plot.zoo'. -) `tick.tstep' and `lab.tstep' now accept the values: "auto", "weeks", "hours", "minutes", "seconds", which should work with sub-daily time series -) when `cal.ini' is provided, all the values in `obs' and `sim' with dates previous to `cal.ini' are SKIPPED from the computation of the goodness-of-fit measures (when 'gof.leg=TRUE'), but their values are still plotted, in order to examine if the warming up period was too short, acceptable or too long for the chosen calibration period. -) fixed problem with the `cex.axis' argument when 'plot.type="multiple"' (this was a zoo problem, and it was fixed in zoo_1.7-1). o `ggof' : -) better display of time series objects (by using the new `plot2' and 'drawxaxis/drawTimeAxis') -) it works with xts objects as well. -) `tick.tstep' and `lab.tstep' now accept the values: "auto", "quarters", "weeks", "hours", "minutes", and "seconds", and its default value was changed to "auto". -) the name of `sim' and `obs' are now used as default for the legend. -) when `cal.ini' is provided, all the values in `obs' and `sim' with dates previous to `cal.ini' are SKIPPED from the computation of the goodness-of-fit measures (when 'gof.leg=TRUE'), but their values are still plotted, in order to examine if the warming up period was too short, acceptable or too long for the chosen calibration period. -) added error checking for `ftype' when `sim' and `obs' are not time series. o `plotbands' : -) better display of time series objects. The (nicer) `plot.xts' function is used instead of the previous `plot.zoo', and it uses the new `drawTimeAxis' function. -) `tick.tstep' and `lab.tstep' now accept the values: "auto", "quarters", "weeks", "hours", "minutes", and "seconds", and its default value was changed to "auto". o `plotbandsonly': -) it works with xts objects as well -) `legend' argument was removed (it was not used) -) `leg.cex' argument was removed (it was not used) o `NSeff' : -) its name was changed to `NSE'. The old name `NSeff' is deprecated but is still working for backwards compatibility. -) now it returns NA instead of an error when the denominator of its formula is equal to zero. o `rNSeff' : -) its name was changed to `rNSE'. The old name `rNSeff' is deprecated but is still working for backwards compatibility. -) now it returns NA instead of an error when the denominator of its formula is equal to zero. o `mNSeff' : -) its name was changed to `mNSE'. The old name `mNSeff' is deprecated but is still working for backwards compatibility. -) now it returns NA instead of an error when the denominator of its formula is equal to zero. o `rd' : -) a warning message is created when -at least one of- the observed values is/are zero, which makes impossible to compute this index (thanks to Tom Purucker !) -) now it returns NA instead of an error when the denominator of its formula is equal to zero. o `rNSE' : -) a warning message is created when -at least one of- the observed values is/are zero, which makes impossible to compute this index -) now it returns NA instead of an error when the denominator of its formula is equal to zero. o `md' : -) now it returns NA instead of an error when the denominator of its formula is equal to zero. o `pbias' : -) now it returns NA instead of an error when the denominator of its formula is equal to zero. o `pbiasfdc': -) now it returns NA instead of an error when the denominator of its formula is equal to zero. o `pfactor' : -) now it returns NA instead of an error when the denominator of its formula is equal to zero. o `rfactor' : -) now it returns NA instead of an error when the denominator of its formula is equal to zero. o `rSD' : -) now it returns NA instead of an error when the denominator of its formula is equal to zero. o `RSR' : -) now it returns NA instead of an error when the denominator of its formula is equal to zero. o `NRMSE' : -) now it returns NA instead of an error when the denominator of its formula is equal to zero. o `KGE' : -) now it returns NA instead of '-Inf' when the denominator of Alpha or Beta were equal to zero. o `rd' : -) fixed documentation error in its formula (thanks to Tom Purucker !). Its value was already computed in the correct way. o New dimension checking (number of columns and rows) for all the functions that are applied over a matrix or data.frame ################################################### # Previous Releases (see old file 'WhatsNew.txt)' # ################################################### 0.2-2: Release Date: 14-Apr-2011 0.2-1: Release Date: 30-Nov-2010 0.2-0: Release Date: 07-Oct-2010 0.1.3: Release Date: 01-Dec-2009 0.1.2: Release Date: 29-Oct-2009 0.1.1: Release Date: 05-Oct-2009 0.1.0: Release Date: 07-Sep-2009 (but the functions have been developed since 2008)