| highNoise {RFPM} | R Documentation |
Simulated Floating Percentile Model Data, High Random Noise in Toxicity Prediction
Description
Example dataset
Format
data.frame, containing 147 rows and 11 variables:
- Al
aluminum concentration, mg/kg dry weight
- As
arsenic concentration, mg/kg dry weight
- Cu
copper concentration, mg/kg dry weight
- Cd
cadmium concentration, mg/kg dry weight
- Cr
chromium concentration, mg/kg dry weight
- Fe
iron concentration, mg/kg dry weight
- Pb
lead concentration, mg/kg dry weight
- Hg
mercury concentration, mg/kg dry weight
- Ni
nickel concentration, mg/kg dry weight
- Zn
zinc concentration, mg/kg dry weight
- Hit
logical; whether the sample was classified as toxic
Details
highNoise provides a sample of simulated data that were developed to analyze variability and sensitivity of FPM.
Simulated datasets (n=1000) were originally developed for the analyses; highNoise is a single realization of the simulation.
Simulations were generated using the covariance matrix of h.northport sediment chemical concentrations and the
rmvnorm function from the 'splus2R' package (Constantine and Hesterberg 2021). Hit values were generated with toxCRM
using simulated concentrations of Cr, Cu, Fe, and Zn
as inputs. The inputs to toxCRM were based on empirical toxicity ranges (minimum and maximum effect levels) as well as
arbitrarily based on probable effect concentration (PEC) and threshold effect concentration (TEC) values from MacDonald et al (2000). Inflection points were
set equal to PECs, and steepnesses were set as the ratio of the PEC to the TEC.
Fe inputs were arbitrary chosen to reflect its low toxicity relative to Cr, Cu, and Zn (e.g., high inflection point and relatively low steepness).
Random normal noise was added to the highNoise datasets by inputting eSD values into toxCRM equal to 20% of the range of chemical concentrations (chemical-specific).
The median toxicity output (for the 4 chemicals) was compared to a threshold of 75% with Hit == TRUE assigned to values below that level and
Hit == FALSE assigned to values >=75%.
References
Constantine W, Hesterberg T. 2021. splus2R: Supplemental S-PLUS Functionality in R. Version 1.3-3 (online). Updated January 30. Available from: https://cran.r-project.org/web/packages/splus2R/index.html. MacDonald DD, Ingersoll CG, Berger TA. 2000. Development and evaluation of consensus-based sediment quality guidelines for freshwater ecosystems. Arch Environ Contam Toxicol 39(5):20-31.
See Also
FPM, toxCRM, h.northport, perfect, lowNoise