Make your results of standard statistical analysis browsable and reproducible by exporting them into JSON-LD, following a standardized vocabulary.
Version: 1.0.3
Dependencies: RJSONIO, formula.tools, stringr, rstan, lmtest
Insert a call to RJSONLD.export for every result you wish to export, along with
the relative path of the file you wish to create:
result = lm(iris$Petal.Length~iris$Sepal.Length)
RJSONLD.export(result,"irisLM.jsonld")will create the following JSON-LD file:
{
"@context" : {
"@vocab" : "http://standardanalytics.io/stats/"
},
"@type" : "LinearModel",
"modelFormula" : "iris$Petal.Length ~ iris$Sepal.Length",
"modelVariable" : [
{
"@type" : "Variable",
"name" : "iris$Petal.Length"
},
{
"@type" : "Variable",
"name" : "iris$Sepal.Length"
}
],
"modelFit" : {
"@type" : "FitnessOptimization",
"fitnessCriterion" : "least squares",
"r2" : 0.75995,
"adjr2" : 0.75833,
"nData" : 150,
"nParameters" : 2,
"fTest" : {
"@type" : "FTest",
"testStatistic" : 468.55,
"dfNum" : 1,
"dfDenom" : 148,
"pValue" : 1.0387e-47
},
"optimalParameter" : [
{
"@type" : "Parameter",
"name" : "(Intercept)",
"estimate" : {
"@type" : "Statistic",
"value" : -7.1014,
"standardError" : 0.50666,
"statisticalTest" : {
"@type" : "TTest",
"testStatistic" : -14.016,
"df" : 148,
"pValue" : 6.1336e-29
}
}
},
{
"@type" : "Parameter",
"name" : "iris$Sepal.Length",
"estimate" : {
"@type" : "Statistic",
"value" : 1.8584,
"standardError" : 0.085856,
"statisticalTest" : {
"@type" : "TTest",
"testStatistic" : 21.646,
"df" : 148,
"pValue" : 1.0387e-47
}
}
}
]
},
"fitResidual" : [
{
"@type" : [
"Statistic",
"Min",
"Quantile"
],
"name" : "Min",
"value" : -2.4775,
"percentile" : 0
},
{
"@type" : [
"Statistic",
"Quantile"
],
"name" : "1Q",
"value" : -0.59072,
"percentile" : 25
},
{
"@type" : [
"Statistic",
"Median",
"Quantile"
],
"name" : "Median",
"value" : -0.0066844,
"percentile" : 50
},
{
"@type" : [
"Statistic",
"Quantile"
],
"name" : "3Q",
"value" : 0.60484,
"percentile" : 75
},
{
"@type" : [
"Statistic",
"Max",
"Quantile"
],
"name" : "Max",
"value" : 2.4951,
"percentile" : 100
}
]
}MIT