The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.
- all examples have been streamlined to shorten the time required for executing R package checks
- added detailed explanations for the necessity of using donttest{}
- PCA option is removed from the daily_response and other similar functions
- plot specific options is also removed
- New climate detrending option is implemented, ‘SLD’, which stands for ‘Simple Linear Detrending’ and should be preferred
- Two new arguments are implemented, ‘skip_window_position’ and ‘skip_window_length’ which allow for controlling the granularity of the analysis and can help in reducing computation time by focusing on a subset of the data
- Spline detrending is now removed since it caused instability of the dendroTools package
- response data frame can now also have a missing value, but only if “cor_na_use” argument is one of “complete.obs”, “na.or.complete” or “pairwise.complete.obs”
- an undocumented argument in internal function is now removed
- A bug in some special cases for the reference window “end” is now removed
- In case of no significant correlations, daily_response() now returns a warning rather than error
- new options for the aggregate function - minimum and maximum
- additional feature on reference_window in the monthly_response() and monthly_response_seascorr()
- daily_response() default remove_insignificant argument is set to FALSE
- updated description
- default method in daily_response() and monthly_response() is set to ‘cor’
- corrected all spelling errors
- comments are added to the code, which is now more readable
- unused part of the scripts are removed
- new arguments are implemented to control the behavior of calculations of correlation coefficients in case of missing values
- the calculation of significance of partial correlation coefficients is improved
- daily_transform() is also improved, so it gives the correct behavior in case of missing values
- detrending climate data is now available in all response type functions
- updated vignettes
- new arguments are added which can be used to detrend climate
- climate detrending is based on the dplR R package
- gridExtra is removed from dependences
- Minor correction for ggplot2 object -> guide = FALSE –> guide = ‘none’
- There is a major upgrade related to new arguments that now allow you to specify the time of interest to be used for calculating correlation coefficients or other statistical metrics. For example, if you are using daily_response() or daily_response_seascorr() and want to calculate correlations only from the beginning of last October to the current end of September, use the argument daily_interval =c(-274, 273). Note that a negative sign indicates the previous year’s doy, while a positive sign indicates the current year doy. In normal (non-leap ) years, October 1 represents doy 274, and September 30 represents doy 273.
- If you use monthly_response() or monthly_response_seascorr() and want to calculate correlations only from the beginning of the previous October to the current end of September, for example, use the argument monthly_interval =c(-10, 9). Note that a negative sign indicates the month of the previous year, while a positive sign indicates the month of the current year. Other functions have been changed only slightly, so they will work even if new arguments are used.
- new error is implemented in case of wrong subset_years
- minor corrections to partial correlation functions
- deleted example in vignette Examples_with_daily_climate_data which caused problems with CRAN check
- a bug corrected in the data_transform()
- minor bug is removed which happened in a case of non significant calculations or only one significant correlation
- dendroTools version is changed to 1.1.1
- Corrections related to testthat changes.
- imports is updated: oce(>= 1.2-0)
- Minor corrections after new publication
- Sys.setlocale(“LC_TIME”, “C”) is now added, so optimal dates will always show in English language
- The vignette titles are changed
- Bootstrapping is now available for all metrics, monthly and daily functions
- Seacorr is now available also for monthly data, including bootstrapping
- swit272 daily precipitation dataset is added
- swit272 monthly temperature dataset is removed
- KNMI_daily_transform() is now called data_transform(). The entire function is updated and now also enables transformation of daily data into monthly.
- Bootstrapping of correlation coefficients is introduced. To do so, use the argument boot in daily_response() and monthly_response(). Bootstrapping is currently available only for correlation coefficients. In future version, it will also be available for model fitting.
- Package version is changed to 1.0.6
- new function: monthly_response_seascorr()
- titles of plot_extreme, plot_specific and plot_heatmap are updated
- for daily_response() and monthly_response(), it is now possible to define method for correlation coefficient: “pearson”, “kendall”, “spearman”. To do so, use , cor_method argument
- daily_response output list is class “dmrs”
- Using newly defined class “dmrs”, new summary function is defined
- Package version is changed to 1.0.5
- New data is available for examples demonstration: swit272_monthly_temperatures
- New function is available: daily_response_seascorr(). This function implements partial correlations in the analysis of daily response functions
- Package version is changed to 1.0.4
- Package version is changed to 1.0.3
- There is new function introduced: KNMI_daily_transform() which transforms daily data obtained from KNMI Climate explorer into data frame suitable for daily_response().
- New TRW chronology is included, swit272
- New function: monthly_response()
- New data: LJ_monthly_temperatures
- New data: LJ_monthly_precipitation
- Package version is changed to 1.0.2
- There have been many issues with RWeka package, which depends on rJava. Therefore, functions from RWeka are replaced with other functions. For random forest model we now use randomForest package
- There are six new output elements, all of them are residual diagnostic plots for calibration, holdout and edge data.
- Due to many problems related to rJava in R, MT and BMT methods from RWeka are now joined and replaced by cubist function from Cubist r package. Cubist fits a model tree or ensemble of model trees, if committees argument is more than 1.
These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.