aquap2 is now making use of the package 'uniset' for handling the settings.R file.
Example datasets / experiments can now be installed easily (function ap2dme).
Possibility to manually denote abberations in number of consecutive scans in the sample list.
Function to manually multiply a single-line sample list (sampleList_multiplyRows)
Sample list file is now exclusively in xlsx format
The Metadata file now contains a variable for the sample list type, the temperature and humidity log file and for multiplying the rows within the sample list. Thus it is possible to encode all required information to import data statically in the metadata file, thus being able to call gfd() without additional parameters.
Added exporting of full dataset to xlsx.
Added importing from xlsx rawdata: Can be either NIR only, or NIR + header.
Added possibility to export only the header to xlsx.
Added import function for Yunosato Aquaphotomics lab styled .dat file.
Added possibility to automatically remove double columns during data import.
Added importing from MicroNIR (VIAVI) data format
Added importing T and RH data from Hoboware data loggers
Improved:
Removed dependency from package ChemometricsWithR (is not available on CRAN any more)
Completely overhauled the creation of the sample list: A xlsx file is used now to define the class and numerical variables.
Improved import when importing from raw-data file AND sample list file. Double columns now simply get deleted.
Bugfix:
Version 0.4.4
New:
In Aquagrams, bootR is now adjusted to the actual dataset.
Improved:
Aquagram
Bugfix:
Various small bugfixes
Version 0.4.2
New:
Create new variables via combining pre-existent values or by calculating a new value using an arbitrary expression.
Improved:
When merging datasets, it is now possible to also resample the wavelengths if necessary.
Bugfix:
Various small bugfixes (e.g. the show-method for aquap_data)
Version 0.4.1
New:
Added merging of datasets.
Improved:
Bugfix:
Fixed a bug that caused the path for the settings home folder in the .Renviron file to be wrong on Windwos machines.
Version 0.3.2
New:
Improved:
Bugfix:
Fixed "stringsAsFactors" problem in creating the header data-frame (due to the update to R 4.x)
Version 0.3.1
New:
option in the settings to use also virtual cores in parallel computing
Improved:
classic aquagram is now impossible to bootstrap (as that does not make any sense)
Bugfix:
fixed the parallel calculation of aquagrams on windows systems
Version 0.2.0
New:
Added classifiers.
Function to select and possibly average groups of wavelengths (function siWlg).
Version 0.1.9
New:
Values on the X-Axis can be others than wavelengths; the denominator can be specified in the metadata.
Bugfix:
Repaired custom color vectors in Aquagrams.
Other small bugfixes.
Version 0.1.8
New:
Added possibility to plot Aquagrams linear instead of circular. Factory-default is linear;
parameter "aqg_plottingType" in the settings file.
Added output of table indicating significant differences between groups and the corresponding values.
Added some more Aquagram specific options to the settings file.
Resampling of data to new wavelengths (core data manipulation function).
Added possibility to average spectra by any class-variable grouping (function "do_avg").
Improved:
If exact temperature is present in temperature calibration data, "loess" is not used any more.
In the "diff" modes of the Aquagram, the input for "aqg.minus" is now vectorized, i.e. you can supply
a character vector in "aqg.minus" for corresponding subtractions in the character vector in "aqg.vars".
Bugfix:
Classic Aquagram can now be produced without having a temperature calibration file.
Outlier-detection using the package rrcovHD should now work on all systems.
Now correct detection of range (precise wavelength independent) in universal AUC table.