README.md

In MikkoVihtakari/veganUtils: Utility functions and extensions to vegan and ggvegan

veganUtils

Utility functions and extensions to vegan. R package version 0.1.9

While the vegan package is R's leading multivariate and community ecology tool, producing illustrative graphics to examine differences in species patterns often requires fine-tuning the original graphics. Such fine tuning often requires long piece of code before vegan's cca or rda results are in a visually examinable format.

The veganUtils package produces good looking and illustrative ordination graphics for vegan with less code. The package also contains other utility functions and extensions to improve the output of vegan objects together with tools to graphically examine nonlinear relationships in multivariate data. Further, many plotting functions in veganUtils are implemented using ggplot2. The package complements the vegan package using solutions from this package whenever possible.

The package is under development and has not been uploaded to CRAN, but veganUtils can be installed directly from GitHub

remotes::install_github("MikkoVihtakari/veganUtils")

The package is in an early developmental phase, and some functions do not work as intended. Note that the package comes with absolutely no warranty. Always check the ordination graphics returned by veganUtils using the vanilla vegan plots. Any bug reports and code fixes are warmly welcomed. Borrowing code from the package to improve vegan and associated packages is encouraged. See Contributions and contact information for further details.

Usage

The intention here is not to give instructions for scientifically valid multivariate analyses, but rather to illustrate the differences between vegan and veganUtils graphics. Refer to the vegan tutorial for instructions how to use vegan correctly.

The description under is now finished and will be improved.

Non-constrained ordination

Load datasets and make a principal analysis (rda) object:

library(vegan)
library(veganUtils)

data(BCI)
data(BCI.env)

x <- decostand(BCI, method = "hellinger")

mod <- rda(x)

Original vegan ordination graphics

plot(mod)

veganUtils ordination graphics alternative using base graphics

ord_plot(mod)

veganUtils ordination graphics alternative using ggplot2

ggord_plot(mod)

Adding envfit and ordisurf

The envfit and ordisurf fits can be added directly using the ord_plot function:

ord_plot(mod, env_data = BCI.env, ordisurf_var = "UTM.EW", envfits = c("Habitat", "UTM.EW"))

Transform envfit objects to clean data frames

The envfit_table function prints a data frame using the vegan´s envfit function. Useful for quickly assessing linear fit of environmental variables to an ordination.

envfit_table(mod, BCI.env)

##     Type      Variable   R2     p   PC1   PC2
## 1 Factor       Habitat 0.40 0.001    NA    NA
## 2 Factor        Stream 0.03 0.252    NA    NA
## 3 Factor       Age.cat 0.01 0.838    NA    NA
## 4 Factor       Geology 0.00 1.000    NA    NA
## 5 Vector        UTM.EW 0.85 0.001 -0.08 -1.00
## 6 Vector        EnvHet 0.08 0.114  0.08 -1.00
## 7 Vector        UTM.NS 0.04 0.395  0.42  0.91
## 8 Vector Precipitation 0.00 1.000  0.00  0.00
## 9 Vector     Elevation 0.00 1.000  0.00  0.00

The function also contains a feature to clean up variable names

envfit_table(mod, BCI.env, clean_var_names = TRUE)

##     Type      Variable   R2     p   PC1   PC2
## 1 Factor       Habitat 0.40 0.001    NA    NA
## 2 Factor        Stream 0.03 0.314    NA    NA
## 3 Factor       Age cat 0.01 0.823    NA    NA
## 4 Factor       Geology 0.00 1.000    NA    NA
## 5 Vector        UTM EW 0.85 0.001 -0.08 -1.00
## 6 Vector        EnvHet 0.08 0.143  0.08 -1.00
## 7 Vector        UTM NS 0.04 0.390  0.42  0.91
## 8 Vector Precipitation 0.00 1.000  0.00  0.00
## 9 Vector     Elevation 0.00 1.000  0.00  0.00

Constrained ordination

mod <- rda(x ~ Habitat + UTM.EW, BCI.env)

Original vegan constrained correspondance analysis ordination graphics

plot(mod)

veganUtils ordination graphics alternative using base graphics. Assigning the explanatory data matrix to env_data cleans up CCA arrow labels automatically.

ord_plot(mod, env_data = BCI.env)

veganUtils ordination graphics alternative using ggplot2

ggord_plot(mod, size_preset = "html")

ord_plot options

ggord_plot options

Axis contributions

axis_plot(mod, italicize_sp_names = TRUE)

Examine nonlinear relationships in multivariate data

pattern_plot(mod)

Contributions and contact information

Any contributions to the package are more than welcome. Please contact the package creator Mikko Vihtakari ([e-mail]) to discuss your ideas on improving the package. Borrowing code is encouraged, but please cite the original sources.

Dependencies

The veganUtils package depends on:

• vegan: veganUtils expands on vegan.
• ggplot2: Used for graphics.
• ggrepel: Used to repel species labels in ggord_plot
• reshape2: Used to manipulate data.
• plyr: Used to access the mapvalues, revalue and round_any functions.
• gridExtra:
• Hmisc: Used to capitalize first letter in strings. Remove.
• tibble: Used to add columns in between columns in data frames. Remove or make dependancy to tidyverse

MikkoVihtakari/veganUtils documentation built on May 3, 2023, 9:58 a.m.