In fmichelsendis/sendis: A Systematic Evaluated Nuclear Data Integral Survey

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>", 
  echo=FALSE, 
  message=FALSE, 
  warning=FALSE, 
  eval=TRUE,
  fig.align='center'
  #fig.width = 100pct
  ) 

htmltools::tagList(rmarkdown::html_dependency_font_awesome())

library(dplyr)
library(plotly)

The sendis package was created with the aim to facilitate the import of new data. In general, if you are into criticality calculations, the first thing you will want to do is to see how other people's results compare to your calculated data. Here are the instructions to do so.

CSV template

You can import your own data provided it is in this csv template. This template is designed to be the ouput of a criticality benchmark calculation : each observation (e.g. one benchmark result per row) has 7 features or variables (in columns). This is the format that the calcs dataset follows. Below is an example of what your csv file structure should look like before it can be imported.

library(sendis)
library(kableExtra)
set.seed(137)
# sample n random entries from calcs : 
df<-calcs
n<-20
df<-df[sample(nrow(df), n), ]
rownames(df)<-NULL
df<-df%>%mutate(CODE="MYCODE",INST="MYLAB")
knitr::kable(df)%>%
  kable_styling() %>%
  scroll_box(width = "100%", height = "200px")

Format and description of variables

FULLID : [character] ICSBEP evaluation identifier, in the format FISS-FORM-SPEC-CASE-SUBCASE;
MODEL : [character] model input used, when specified, otherwise leave blank;
CALCVAL : [numerical], calculated value of $k_{e!f!f}$;
CALCERR : [numerical], statistical uncertainty on $k_{e!f!f}$;
INST : [character] used to identify institute or author of calculation
CODE : [character] name of the code used;
LIBVER : [character] library used, in the format LIB-VER (e.g.: JEFF-3.3)

Note the format for the two categorical values FULLID and LIBVER is extremely important, as these columns are parsed and split at the hyphen ('-') into separate variables.

Importing into `sendis`

The function sendis::import_submissions() imports all *.csv files in the directory specified as its argument path. This function will fail with an error if your csv files are not in the csv template that is required.

# use function import_submission(path="yourfolder")
df<-import_submissions(path = "../docs/files/mycsvfiles/")

# verify import by checking the dimensions of dataframe `mydata`: 
dim(df)

# dataframe contains all of your csv data in `yourfolder/` :
head(df)

Your data is imported but not yet in the format of the sendis dataframe. You only need to call function compile_to_sendis() in order to do so :

# merge your dataframe and bind it to the sendis dataframe
mydata<-compile_to_sendis(df)

# verify import by checking the new dimensions of dataframe `mydata`: 
dim(mydata)
dim(sendis)

Try plotting it :