leppott / xc95 Goto Github PK

Several issues.

1. Not generic enough. Some variables still hard coded into the function (e.g., always wants 3 groups and they are HUC, EcoRegion, and WS Area).

2. Function name should lose the "fish" name. Will keep fish.wt.cdf but will hide it.

3. Includes a subdirectory.

4. Not every parameter should be defaulted. The data files especially.

The option to add a map in the tolerance function only works if the maps library is in the global environment.

1. Update the function to use maps::map

2. Add the maps library to the DESCRIPTION file.

Add a boot strap function/routine.

Not in current functions.

nb

The function auc() (area under curve) appears in multiple functions. Should pull out into its own function.

nb

Need to define, in the help file, the fields in the output.

names(cont.opt)
[1] "tnames" "N" "min_ob" "X50_th_ob" "X95_th_ob"
[6] "max_ob" "Opt_WA" "Tol_WA" "CDF_50_th_Abund" "CDF_95_th_Abund"
[11] "CDF_50_th_PA" "CDF_95_th_PA" "CDF_wt_50_th" "CDF_wt_95_th" "LRM_50_th"
[16] "LRM_95_th" "QLRM_50_th" "QLRM_95_th" "Opt_qlrm" "Tol_qlrm"
[21] "GAM_50_th" "GAM_95_th" "ROC"

Is your feature request related to a problem? Please describe.
Not a separate function. Basic calculation.

The HC05 is the 5th percentile hazardous concentration of a set of related 95th percentile extirpation concentration values (XC95). The HC05 is the estimated value of the stressor at which 5% of species will be extirpated.

Describe the solution you'd like
Pulled from BC paper code, March 2017 and some other files from June 2016.

Describe alternatives you've considered
Add as an example

Additional context
Run example for fish.wt.cdf to get XC95 values.

# data
data(dta.do)
data(ss.sites)
# run function (~20 seconds)
dftv.do <- fish.wt.cdf(datafile = dta.do, ss = ss.sites, plot = T, dogam = T
                      , SampleID = "Station_Date", tag = "wt", sortvect = NULL
                      , np = 61, nt = 25, addtrend = T
                      , wd = getwd(), groups = c("BigHUC","ECOREGL3","WS_AREA")
                      , xvar = "cond")
View(dftv.do)

knitr::kable(head(dftv.do))

Then calculate the 5th quantile.

hc05_cdf <- quantile(dftv.do$XC95.cdf, probs = 0.05, na.rm = TRUE, type = 6)
hc05_gam <- quantile(dftv.do$XC95.gam, probs = 0.05, na.rm = TRUE, type = 6)
# both at once
hc05_both <- apply(dftv.do[, c("XC95.cdf", "XC95.gam")], 2
    , quantile, probs=0.05, na.rm=TRUE, type=6)

Describe the bug
Tibbles sometimes behave different from data frames.

For example, if use read_csv rather than read.csv the user creates a tibble. The way tibbles react when subsetted can throw off expected data formats inside of the functions.

https://www.r-bloggers.com/using-r-when-weird-errors-occur-in-packages-that-used-to-work-check-that-youre-not-feeding-them-a-tibble/

To Reproduce
Steps to reproduce the behavior:

1. read a files with read_csv (or similar function) that creates a tibble.
2. Most of the functions will fail with mostly non-easy to interpret errors.

Expected behavior
A clear and concise description of what you expected to happen.

Screenshots
If applicable, add screenshots to help explain your problem.

Additional context
Can fix (and have in other packages) by adding a line to convert any input that might be a tibble to a dataframe.

df.x <- as.data.frame(input.df.x)

nb

Need to post to CRAN.