static double calcSDev (REngine engine, double[] arr)
{
// Note: only one quick and slightly dirty way to do it
NumericVector rVector = engine.CreateNumericVector(arr);
engine.SetSymbol ("x", rVector);
return engine.Evaluate ("sd(x)").AsNumeric () [0];
}
/// <summary>
///
/// </summary>
/// <param name="data"></param>
/// <returns></returns>
public NumericVector RNumeric(double data)
{
try
{
NumericVector vector = engine.CreateNumericVector(new[] { data });
vector.Protect();
return(vector);
}
catch (SEHException e)
{
throw new ApplicationException("R Engine has thrown an Exception - memory possibly corrupted; Vector could not be created in R.", e);
}
finally
{
ClearMemory();
}
}
public static void CreateNumericVector(REngine engine, int n, Stopwatch s)
{
double[] d = createDoubleArray(n);
s.Start();
var nvec = engine.CreateNumericVector(d);
s.Stop();
}
static void TestOptimCsharp(REngine engine)
{
var rand = new Random(0);
int n = 10000;
double x, y, r, xb, yb, rb;
rb = double.MaxValue; xb = yb = double.MaxValue;
engine.Evaluate("rosen <- function(x, y) { (1-x)**2 + 100*(y-x*x)**2 }");
Console.WriteLine("*** Try a basic way to call the function in R ***");
for (int i = 0; i < n; i++)
{
x = -1 + rand.NextDouble() * (3 - (-1));
y = -1 + rand.NextDouble() * (3 - (-1));
r = engine.Evaluate(string.Format("rosen({0}, {1})", x, y)).AsNumeric().ToArray()[0];
if (r < rb)
{
rb = r;
xb = x;
yb = y;
}
}
Console.WriteLine("The best score r={0} is for x={1}, y={2}", rb, xb, yb);
Console.WriteLine("*** Try an R function 'pointer' with a vectorized function call. Faster, if you can do it this way***");
var f = engine.GetSymbol("rosen").AsFunction();
double[] xa = new double[n], ya = new double[n];
rand = new Random(0);
for (int i = 0; i < n; i++)
{
xa[i] = -1 + rand.NextDouble() * (3 - (-1));
ya[i] = -1 + rand.NextDouble() * (3 - (-1));
}
double[] ra = f.Invoke(new[] { engine.CreateNumericVector(xa), engine.CreateNumericVector(ya) })
.AsNumeric().ToArray();
rb = ra.Min();
int indBest = -1;
for (int i = 0; i < ra.Length; i++)
{ // no which.min in C#. Should call R here too...
if (ra[i] <= rb)
indBest = i;
}
Console.WriteLine("The best score r={0} is for x={1}, y={2}", rb, xa[indBest], ya[indBest]);
}
internal static SymbolicExpression ToVector(REngine engine, IEnumerable values)
{
if (values == null)
{
throw new ArgumentNullException("values", "values to transform to an R vector must not be null");
}
var ints = values as IEnumerable <int>;
var chars = values as IEnumerable <string>;
var cplxs = values as IEnumerable <Complex>;
var logicals = values as IEnumerable <bool>;
var nums = values as IEnumerable <double>;
var raws = values as IEnumerable <byte>;
var sexpVec = values as SymbolicExpression;
if (sexpVec != null && sexpVec.IsVector())
{
return(sexpVec);
}
if (ints != null)
{
return(engine.CreateIntegerVector(ints));
}
if (chars != null)
{
return(engine.CreateCharacterVector(chars));
}
if (cplxs != null)
{
return(engine.CreateComplexVector(cplxs));
}
if (logicals != null)
{
return(engine.CreateLogicalVector(logicals));
}
if (nums != null)
{
return(engine.CreateNumericVector(nums));
}
if (raws != null)
{
return(engine.CreateRawVector(raws));
}
throw new NotSupportedException(string.Format("Cannot convert type {0} to an R vector", values.GetType()));
}
private static DataFrame UserReported(REngine engine)
{
// Incomplete data and repro info.
// See https://rdotnet.codeplex.com/discussions/569196
NumericVector PreprocessedValue = null;
DataFrame PredictedData = null;
// Some info was missing. Make up.
string StartDate = "2001-01-01";
double[] PreProcessedList = new[] { 1.1, 7.3, 4.5, 7.4, 11.23, 985.44 };
string days = "2"; string interval = "3";
PreprocessedValue = engine.CreateNumericVector(PreProcessedList);
// Assign the Utilization value to R variable UtilValue
engine.SetSymbol("PreprocessedValue", PreprocessedValue);
engine.Evaluate("library(forecast)");
engine.Evaluate("StartDate <- as.Date('" + StartDate + "') + " + days);
engine.Evaluate("size = length(seq(from=as.Date('" + StartDate + "'), by='" + "day" + "', to=as.Date(StartDate)))");
engine.Evaluate("startDate <- as.POSIXct('" + StartDate + "')");
engine.Evaluate("endDate <- StartDate + as.difftime(size, units='" + "days" + "')");
engine.Evaluate("PredictDate = seq(from=StartDate, by=" + interval + "*60, to=endDate)");
engine.Evaluate("freq <- ts(PreprocessedValue, frequency = 20)");
engine.Evaluate("forecastnavie <-snaive(freq, Datapoints)");
engine.Evaluate("PredictValue = (forecastnavie$mean)");
engine.Evaluate("PredictedData = cbind(PredictValue, data.frame(PredictDate))");
PredictedData = engine.Evaluate("PredictedData").AsDataFrame();
return PredictedData;
}
private static void ReproWorkitem43(REngine engine)
{
Random r = new Random(0);
int N = 500;
int n1 = 207;
int n2 = 623;
var arGroup1Intensities = new double[N][];
var arGroup2Intensities = new double[N][];
for (int i = 0; i < N; i++)
{
arGroup1Intensities[i] = new double[n1];
arGroup2Intensities[i] = new double[n2];
for (int j = 0; j < n1; j++)
arGroup1Intensities[i][j] = r.NextDouble();
for (int j = 0; j < n2; j++)
arGroup2Intensities[i][j] = r.NextDouble();
}
var res = new GenericVector[N];
NumericVector vGroup1, vGroup2;
for (int i = 0; i < N; i++)
{
vGroup1 = engine.CreateNumericVector(arGroup1Intensities[i]);
Console.WriteLine(vGroup1.Length);
if (i % 10 == 4)
{
engine.ForceGarbageCollection();
engine.ForceGarbageCollection();
}
vGroup2 = engine.CreateNumericVector(arGroup2Intensities[i]);
Console.WriteLine(vGroup2.Length);
engine.SetSymbol("group1", vGroup1);
engine.SetSymbol("group2", vGroup2);
GenericVector testResult = engine.Evaluate("t.test(group1, group2)").AsList();
res[i] = testResult;
}
}
public static void CreateNumericVector(REngine engine, int n, Stopwatch s)
{
double[] d = createDoubleArray(n);
s.Start();
var nvec = engine.CreateNumericVector(d);
s.Stop();
}
internal static SymbolicExpression ToVector(REngine engine, IEnumerable values)
{
if (values == null) throw new ArgumentNullException("values", "values to transform to an R vector must not be null");
var ints = values as IEnumerable<int>;
var chars = values as IEnumerable<string>;
var cplxs = values as IEnumerable<Complex>;
var logicals = values as IEnumerable<bool>;
var nums = values as IEnumerable<double>;
var raws = values as IEnumerable<byte>;
var sexpVec = values as SymbolicExpression;
if (sexpVec != null && sexpVec.IsVector())
return sexpVec;
if (ints != null)
return engine.CreateIntegerVector(ints);
if (chars != null)
return engine.CreateCharacterVector(chars);
if (cplxs != null)
return engine.CreateComplexVector(cplxs);
if (logicals != null)
return engine.CreateLogicalVector(logicals);
if (nums != null)
return engine.CreateNumericVector(nums);
if (raws != null)
return engine.CreateRawVector(raws);
throw new NotSupportedException(string.Format("Cannot convert type {0} to an R vector", values.GetType()));
}
static void Main(string[] args)
{
SetupPath(); // current process, soon to be deprecated
// There are several options to initialize the engine, but by default the following suffice:
REngine engine = REngine.GetInstance();
engine.Initialize(); // required since v1.5
// some random weight samples
double[] weight = new double[] { 3.2, 3.6, 3.2, 1.7, 0.8, 2.9, 2, 1.4, 1.2, 2.1, 2.5, 3.9, 3.7, 2.4, 1.5, 0.9, 2.5, 1.7, 2.8, 2.1, 1.2 };
double[] lenght = new double[] { 2, 3, 3.2, 4.7, 5.8, 3.9, 2, 8.4, 5.2, 4.1, 2.5, 3.9, 5, 2.4, 3.5, 0.9, 2.5, 2.7, 2.8, 2.1, 1.2 };
// introduce the samples into R
engine.SetSymbol("weight", engine.CreateNumericVector(weight));
engine.SetSymbol("lenght", engine.CreateNumericVector(lenght));
// set the weights and lenghts as a data frame (regular R syntax in string)
engine.Evaluate("df <- data.frame(id=c(1:length(weight)), weight = weight,lenght = lenght )");
// evaluate and retrieve mean
double avg = engine.Evaluate("mean(df$weight)").AsNumeric().ToArray()[0];
// same for standard deviation
double std = engine.Evaluate("sd(df$weight)").AsNumeric().ToArray()[0];
// NumericVector coeff = engine.Evaluate("coefficients(lm(df$weight ~ df$lenght ))").AsNumeric();
// print output in console
System.Globalization.CultureInfo ci = new System.Globalization.CultureInfo("en-gb");
//Show in console the weight and lenght data
Console.WriteLine(string.Format("Weights: ({0})", string.Join(",",
weight.Select(f => f.ToString(ci)) // LINQ expression
)));
Console.WriteLine(string.Format("Length: ({0})", string.Join(",",
lenght.Select(f => f.ToString(ci)) // LINQ expression
)));
Console.WriteLine(string.Format("Sample size: {0}", weight.Length));
Console.WriteLine(string.Format(ci, "Average: {0:0.00}", avg));
Console.WriteLine(string.Format(ci, "Standard deviation: {0:0.00}", std));
var result = engine.Evaluate("lm(df$weight ~ df$lenght)");
engine.SetSymbol("result", result);
var coefficients = result.AsList()["coefficients"].AsNumeric().ToList();
double r2 = engine.Evaluate("summary(result)").AsList()["r.squared"].AsNumeric().ToList()[0];
double intercept = coefficients[0];
double slope = coefficients[1];
Console.WriteLine("Intercept:" + intercept.ToString());
Console.WriteLine("slope:" + slope);
Console.WriteLine("r2:" + r2);
string fileName = "myplot.png";
CharacterVector fileNameVector = engine.CreateCharacterVector(new[] { fileName });
engine.SetSymbol("fileName", fileNameVector);
engine.Evaluate("png(filename=fileName, width=6, height=6, units='in', res=100)");
engine.Evaluate("reg <- lm(df$weight ~ df$lenght)");
engine.Evaluate("plot(df$weight ~ df$lenght)");
engine.Evaluate("abline(reg)");
engine.Evaluate("dev.off()");
//The file will save in debug directory
Application.Run(new Form1());
// After disposing of the engine, you cannot reinitialize nor reuse it
engine.Dispose();
}