Back to Questions
2.6k
views2
comments
Using F# and R for plotting, and statistical simulation
f#

I had a simple goal to use R for plotting and simulation (and all other great functions R has).

Below are the steps i took. I am not sure its the correct way but it does the trick great.

1. Download and install RAndFriends from [link:rcom.univie.ac.at]

2.Create a VS project and add as a reference all the files you can find in the COM tab that are from either in  C:\Program Files\statconn\DCOM\tlb or in  C:\Program Files\statconn\DCOM\bin ( references all start by the letter s)

3. Write some code like:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

#light

open System


open STATCONNECTORSRVLib

let R = StatConnectorClass()


R.Init("R")


Console.WriteLine(R.GetConnectorInformation)


R.Close()

4. Buld this and go to the bin directory of your project  there will be a bunch of dll with names of the form Interop.*dll, these are the dll you need to reference in your code for interactive use.

5 . Have some fun with the dlls in the interactive window:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145

#r @"C:\Users\Joachim\Documents\Visual Studio 2010\Calling R\Interop.StatConnControls.dll"


#r @"C:\Users\Joachim\Documents\Visual Studio 2010\Calling R\Interop.STATCONNECTORCLNTLib.dll"


#r @"C:\Users\Joachim\Documents\Visual Studio 2010\Calling R\Interop.StatConnectorCommonLib.dll"


#r @"C:\Users\Joachim\Documents\Visual Studio 2010\Calling R\Interop.STATCONNECTORSRVLib.dll"


open System


open StatConnectorCommonLib


open STATCONNECTORSRVLib

let R = StatConnectorClass()


R.Init("R")


//Statistics


let simNorm  (n:int) =


    

    R.EvaluateNoReturn("simNorm<- rnorm("+string n+")")


    

    unbox<float array> (R.GetSymbol("simNorm"))

let lm y X =


    

    R.SetSymbol("X",X)


    

    R.SetSymbol("y",y)


    

    unbox<float array> (R.Evaluate("lm(\"y~X\")$coef"))

//Plots


let hist data =


    

    R.SetSymbol("d",data)


    

    R.EvaluateNoReturn("hist(d)")

let plot x y t =


    

    R.SetSymbol("x",x)


    

    R.SetSymbol("y",y)


    

    R.EvaluateNoReturn("plot(x,y,type="+t+")")


   


let abline intercept slope =


    

    R.SetSymbol("intercept",intercept)


    

    R.SetSymbol("slope",slope)


    

    R.EvaluateNoReturn("abline(intercept,slope)")


let lines x y t =


    

    R.SetSymbol("x",x)


    

    R.SetSymbol("y",y)


    

    R.EvaluateNoReturn("lines(x,y,type="+t+")")


   


let X = simNorm 1000


let y = simNorm 1000

plot X y ""

let regCoef = lm y X

abline  regCoef.[0] regCoef.[1]

R.Close()

-----------------
Ideally if you have done this in any other way plesae share thanks! Also does anybody understand what is going underneath the hood ?

Like when I set a symbol in R does the program duplicate the data ?

ie 

1
2
3
4

let x = [|1..100|]


R.SetSymbol("xInR",x)

Does the variable xInR reference the same memory as the x  of the F# code ?

.

what's the role of F# in your example?

Any reason not using R directly?

By on 3/14/2010 11:48 PM ()

Not really any reason for the above.

One configuration for using F# and R is to use F# to gather and process the data (maybe connect to database, webcrawling etc.. ), and R as a tool to look at output (histograms, plots).

For sure the data processing can be done in R, but doing it in F# often leads to more readable code (which is key because you don't want to mess this step up ), and it goes really really fast.

By on 3/15/2010 3:54 PM ()

    Topic tags

    IntelliFactory Offices Copyright (c) 2011-2012 IntelliFactory. All rights reserved.
    Home | Products | Consulting | Trainings | Blogs | Jobs | Contact Us | Terms of Use | Privacy Policy | Cookie Policy     		Built with WebSharper