0 vector result in R after running function

Question

After I finished running my function, I kept on getting 0s as my answer:

 niv_density <- function(returns, mu, delta, alpha, beta, t)
{
    t <- 1/t
    gamma <- sqrt(alpha^2 - beta^2)
    result <- rep(0, (1/t))
    for(i in 1:(1/t))
    {
        term3 <- exp(delta*gamma*t + beta*(returns[i] - mu*t))
        term1 <- alpha*delta*t/pi
        term2_1 <- besselY(alpha*sqrt(delta^2*t^2 + (returns[i] - mu*t)^2), 1)
        term2_2 <- sqrt(delta^2*t^2 + (returns[i] - mu*t)^2)
        term2 <- term2_1/term2_2
        result[i] <- (term1*term2*term3)
    }
}

niv_density(returns, 0, 2, 50, 0, 10)
result

After executing the last part, I get a vector of 0s. I think I'm having a problem with global vs. local scopes, but I'm not sure how to fix it.

dput(returns)
structure(c(-0.003859212, 0.011873365, -0.004826217, -0.004006846, 
-0.004527209, -0.005597606, -0.001446292, 0.004890173, 0.001260653, 
-0.005469839, 0.001715495, 0.00776223, -6.79514e-05, -0.002405413, 
-0.00344927, 0.013203733, 0.009007395, -0.002918161, -0.000682757, 
0.003600917, -0.001584568, 0.001778635, 0.003881849, -0.003228443, 
0.00809855, -0.003407655, 0.006570117, -0.001629285, -0.001479157, 
-0.000683758, 0.007489741, 0.007807378, 0.001399056, -0.000578823, 
-0.002437511, -0.000593349, -0.004020762, 0.004744014, -0.001815119, 
0.007757796, -0.002401808, -0.00225831, -0.005162853, -0.002256747, 
0.032891919, 0.005882631, -0.011822966, -0.005744899, -0.004359233, 
0.00405189, 0.017035644, 0.001079738, 0.001845759, -0.004758891, 
0.006067706, -0.006027932, -0.00224155, -0.010844493, 0, -0.003861616, 
-0.004698823, 0.000397524, 0.001840917, 0.013599978, -0.008376557, 
1.92494e-05, 0.010797502, -0.004105023, 0.003119424, -0.004797368, 
-0.001962367, 0.002663974, 0.008489008, 0.007827146, -0.000566674, 
-0.003404669, -0.000160508, -0.003953786, -0.000635631, 0.0023086, 
0.008931147, -0.002761431, 0.013046559, -0.009673618, 0.007572105, 
-0.011309217, 0.003777911, -0.004767721, -0.004096769, 0.003915212, 
-0.005571037, 0.008566323, -0.009063831, -0.011191246, -0.000639167, 
0.002834983, -0.009156367, 0.00189252, 0.007166451, -0.001788182, 
-0.002437146, 0.00226261, -0.010459432, -0.001511577, 0.00039628, 
-0.00349739, 0.009561965, 0.063504871, 0.003492974, 0.009233691, 
0.004795333, -0.003995969, -0.002552804, 6.81834e-05, 0.006134657, 
0.006713932, -0.006875273, -0.005108732, 0.006239377, 0.002293386, 
-0.01121192, -0.005666844, 0.000894577, -0.012511724, 0.00351622, 
-0.009671627, -0.004480382, 0.007385228, -0.009143379, 0.005467177, 
0.017094141, 0.005918621, 0.001514995, -0.001356959, 0.015656296, 
0.001101646, 0.001457523, 0.0051402, -0.005516804, 0.002832519, 
-0.002196811, -0.007752963, 0.009050809, 0.006380147, 0.001995102, 
0.002319077, -0.001788715, 0.000845096, -0.009821598, 0.012634302, 
-0.001457121, 0.000582262, -0.004083585, -0.004021717, -0.000571503, 
0.006159289, -0.010822168, -0.015789222, -0.000657867, 0.013935285, 
0.001312777, -0.001172312, 0.003031039, 0.002482838, -0.010634785, 
0.014015267, 0.005435065, -0.034817949, 0.005145224, -0.007217488, 
0.00458109, 0.012581199, 0.001853981, 0.002118571, -0.011151137, 
-0.007933775, 0.011336262, 0.018212375, 0.007815775, 0.006103632, 
-0.007270438, -0.001066825, 0.001892988, -0.009740379, 0.012057142, 
0.00024459, -0.003702988, 0.014628744, -0.001902607, -7.49322e-05, 
-0.005903797, -0.002481339, -0.004266069, 0.01150386, -0.019888508, 
0.007657512, -0.004649027, 0, 0.002523089, -0.00072238, -0.021153782, 
-0.007969763, 0.005775428, -0.010897333, 0.007468107, -0.009508927, 
0.000464995, -0.002430182, 0.010796022, 0.008898853, -0.013079549, 
0.027112561, -0.015413991, -0.007630787, 0.007033724, -0.017738864, 
-0.015961032, -0.015579591, -0.011802317, -0.002187586, 0.003065715, 
0.013389559, -0.000885034, -0.013701533, 0.001976838, 0.001041955, 
-0.003616062, 0.005344799, 0.007148373, -0.002877552, -0.007681476, 
0.021591165, 0.017966863, -0.058771073, -0.019551973, 0.005203616, 
0.002169669, 0.003884158, -0.022568915, 0.002769004, -0.007779571, 
0.018998803, -0.001212088, 0.002446011, 0.007740844, 0.012532807, 
0.006287039, 0.003958813, 0.01407559, 0.001064047, -0.00862106, 
-0.012296938, -0.013967015, 0.010524923, -0.010789529, 0.011953286, 
0.000738662, -0.016492003, -0.00257709, -0.015437029, 0.004315983, 
0.023337948, 0.008138125, 0.005972748, 0.005915635, 0.010493804, 
-0.011895336, -0.005245454, 0.007409717, 0.012596218, -0.005221382, 
-0.005462129, 0.008785043, 0.009134618, 0.015541224, 0.016072839, 
-0.003827797, 0.000403703, 0.03749696, -0.003386946, -0.008627298, 
-0.030790478, -0.003861794, -0.011426323, 0.001393173, 0.008541783, 
0.009361445, -0.023851831, 0.024814864, -0.019724128, 0.002621807, 
-0.017904622, -0.003584294, -0.019299804, -0.00234839, -0.002685042, 
0.002685042, 0.016590137, 0.001401377, -0.006120481, 0.006690448, 
-0.004740457, -0.005027981, 0.013204038, -0.002742491, 0.005110009, 
-0.006393429, 0.00464228, -0.00270551, -0.011552836, 0.003074876, 
0.005139878, 0.002032361, 0.007603533, 0.010491222, 0.000658875, 
0.003909991, 0.00236732, 0.019192366, -0.00361624, 0.005696264, 
-0.005852811, 0.014805765, 0.00313454, 0.006385073, -0.005475311, 
-0.009195918, 0.008472618, -0.000559148, -0.007272851, 0.003748203, 
0.001156269, 0.004328552, -0.006107929, -0.012121056, 0.002812434, 
-0.009577213, 0.005689626, -0.001941957, 0.006145673, -0.002275509, 
-0.006578825, -0.005345298, -0.000327811, 0.003751791, 0.005053343, 
0.005157952, -0.022100394, -0.007461083, 0.003576376, 0.00093598, 
0.006738706, 0.006976768, 0.001078282, -0.006256189, 0.003313743, 
-0.005955287, 0.011771523, 0.001644383, -0.003459295, 0.032863111, 
-0.007369908, -0.001099451, 0.004745151, 0.012094786, 0.001167328, 
-0.00404787, -0.004345022, -0.001121192, 0.004333763, -0.008483142, 
-0.001578184, -0.00046999, 0.005079249, -0.005970832, 0.005543307, 
0.006722626, 0, 0.001715197, 0.011776868, 0.013308783, -0.004160112, 
-0.000304697, 0.014924613, 0.007204855, -0.00509816, 0.007186504, 
0.002287253, -0.009948655, -0.001000861, -0.00431929, -0.00347645, 
0.005015994, -0.007540969, 0.00558486, -0.005661924, -0.006602168, 
-0.002824197, 0.001939661, 0.006563001, -0.009757559, -0.00978824, 
-0.001247868, 0.002622219, -0.009097288, -0.014394158, -0.00292424, 
0.002644891, -0.005572549, -0.003181826, 0.002676673, 0.007032888, 
0.002127581, 0.005281961, 0.016021024, 0.001232531, 0.005515082, 
0.000450254, 0.003568462, 0.006277841, -0.003823264, -0.032527132, 
0.021873831, -0.003231721, -0.000368515, -0.001397511, -0.010973353, 
-0.011563657, -0.010061858, 0.005714484, 0.007472816, 0.003407539, 
-0.000612977, -0.000800283, -0.001900635, -0.000865432, -0.003630001, 
0.00562073, 0.001858425, 0.010064273, -0.006584881, -0.001470899, 
0.005433816, -0.002510864, -0.001071656, -0.005130965, 2.35065e-05, 
0.003445676, 0.01374472, -0.001123534, 0.006067276, 0.004050843, 
-0.000773321, -0.003401186, 0.001908336, -0.003562041, -0.001180884, 
-0.003133416, 0.005819655, -0.002096198, -4.92007e-05, 0.002838133, 
-0.010010669, 0.00557654, -0.000122526, 0.022760252, -0.005618111, 
0.014434193, 0.001716112, 0.01567573, 0.001566116, -0.003071945, 
-0.018146189, -0.012123038, -0.007480614, 0.007735601, -0.00436506, 
0.003091618, 0.004704796, 0.001184206, 0.010066361, 0.005389096, 
-0.007021784, -0.004211278, -0.001740557, -0.00628043, 0.002434464, 
-0.000333944, 0.010815674, 0.016910153, 0, -0.01318228, -0.002858256, 
0.024721185, 0.001006412, -0.003651077, 0.009682259, -0.007093437, 
-0.002005597, 0.002424598, -0.015024047, 0.015051995, 0.004720944
), na.action = structure(504L, class = "omit"))

Are you attempting to modify 'result' in the global environment from within the function? — Matthew Lundberg, Nov 03 '12 at 20:33

flodel · Accepted Answer · 2012-11-03T20:39:42.353

Your version is not working because result is local to the function's body and it is lost as you exit the function. I suspect you have another result object in your global environment (a vector of zeroes) and that's what you always get when you try to check your result.

Instead, make your function explicitly return result by adding a return statement at the end:

niv_density <- function(returns, mu, delta, alpha, beta, n) {
    t <- 1/n
    gamma <- sqrt(alpha^2 - beta^2)
    result <- rep(0, n)
    for(i in seq_len(n)) {
        term3   <- exp(delta*gamma*t + beta*(returns[i] - mu*t))
        term1   <- alpha*delta*t/pi
        term2_1 <- besselY(alpha*sqrt(delta^2*t^2 + (returns[i] - mu*t)^2), 1)
        term2_2 <- sqrt(delta^2*t^2 + (returns[i] - mu*t)^2)
        term2   <- term2_1/term2_2
        result[i] <- (term1*term2*term3)
    }
    return(result)
}

And when calling the function, assign the result as follows:

result <- niv_density(returns, 0, 2, 50, 0, 10)

(and maybe you should avoid calling a variable result, I'm sure you can find a more descriptive name from the context.)

Since R returns the last line, you can make do with just `result` at the end. `return` is really only necessary for early function exits. — seancarmody, Nov 03 '12 at 20:43
@seancarmody, yes both work and it is really a matter of taste, see http://stackoverflow.com/questions/11738823/explicitly-calling-return-in-a-function-or-not for an interesting discussion on that topic. — flodel, Nov 03 '12 at 20:45
I saw that when it was first asked and saw @Alan's answer, but hadn't seen the later discussion (including your own), so thanks for the re-point! — seancarmody, Nov 03 '12 at 20:51

0 vector result in R after running function

1 Answers1