Looping through class ltraj? - r
Apologies if this is not the best forum to inquire this question.
Has anyone been able to loop/iterate though multiple GPS collared individuals that are in class ltraj (adehabitatlt)?
I've been trying to calculate Prox (https://cran.r-project.org/web/packages/wildlifeDI/vignettes/wildlifeDI-vignette.pdf) for multiple individuals but am struggling with how exactly to loop class ltraj because it's different than dataframe (which I'm accustomed to).
Thanks in advance.
install.packages('wildlifeDI', dependencies=TRUE)
library(wildlifeDI)
library(adehabitatLT)
chupacabra <- structure(list(CollarID = structure(c(1L, 1L, 1L, 1L, 1L, 1L,
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L,
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L,
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L,
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L,
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L,
1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L,
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L,
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L,
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L,
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L,
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 3L, 3L, 3L, 3L,
3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L,
3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L,
3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L,
3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L,
3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L,
3L, 3L, 3L, 3L, 3L), .Label = c("A4116F", "A4117M", "A4118F"), class = "factor"),
DateTime = structure(c(1433653200, 1433667600, 1433682060,
1433682300, 1433682600, 1433682900, 1433683200, 1433683500,
1433683800, 1433684100, 1433684400, 1433684700, 1433685000,
1433685300, 1433685600, 1433685900, 1433686200, 1433686500,
1433686800, 1433687100, 1433687400, 1433687700, 1433688000,
1433688300, 1433688600, 1433688900, 1433689200, 1433689500,
1433689800, 1433690100, 1433690400, 1433690700, 1433691000,
1433691300, 1433691600, 1433691900, 1433692200, 1433692500,
1433692800, 1433693100, 1433693400, 1433693700, 1433694000,
1433694300, 1433694600, 1433694900, 1433695200, 1433695500,
1433695800, 1433696100, 1433696400, 1433710860, 1433714400,
1433714700, 1433715000, 1433715300, 1433715600, 1433715900,
1433716200, 1433716500, 1433716800, 1433717100, 1433717400,
1433717700, 1433718000, 1433718300, 1433718600, 1433718900,
1433719200, 1433719500, 1433719800, 1433720100, 1433720400,
1433720700, 1433721000, 1433721300, 1433721600, 1433721900,
1433722200, 1433722500, 1433722800, 1433723100, 1433723400,
1433723700, 1433724060, 1433724300, 1433724600, 1433724900,
1433725200, 1433653200, 1433667660, 1433682060, 1433682300,
1433682600, 1433682900, 1433683200, 1433683500, 1433683800,
1433684100, 1433684400, 1433684700, 1433685000, 1433685300,
1433685660, 1433685900, 1433686200, 1433686500, 1433686800,
1433687100, 1433687400, 1433687700, 1433688000, 1433688300,
1433688660, 1433688900, 1433689200, 1433689500, 1433689800,
1433690100, 1433690400, 1433690700, 1433691000, 1433691300,
1433691600, 1433691900, 1433692200, 1433692500, 1433692800,
1433693100, 1433693400, 1433693700, 1433694060, 1433694300,
1433694600, 1433694900, 1433695200, 1433695500, 1433695800,
1433696100, 1433696400, 1433710860, 1433714400, 1433714700,
1433715000, 1433715300, 1433715600, 1433715900, 1433716200,
1433716500, 1433716800, 1433717100, 1433717400, 1433717700,
1433718000, 1433718300, 1433718600, 1433718900, 1433719200,
1433719500, 1433719800, 1433720100, 1433720400, 1433720700,
1433721000, 1433721300, 1433721600, 1433721900, 1433722200,
1433722500, 1433722800, 1433723100, 1433723400, 1433723700,
1433724000, 1433724300, 1433724600, 1433724900, 1433725200,
1433653249, 1433667666, 1433682089, 1433682349, 1433682632,
1433682936, 1433683234, 1433683536, 1433683837, 1433684144,
1433684443, 1433684739, 1433685031, 1433685370, 1433685634,
1433685935, 1433686236, 1433686536, 1433686826, 1433687142,
1433687448, 1433687736, 1433688034, 1433688337, 1433688649,
1433688936, 1433689236, 1433689531, 1433689827, 1433690139,
1433690433, 1433690736, 1433691048, 1433691336, 1433691634,
1433691941, 1433692236, 1433692535, 1433692833, 1433693129,
1433693434, 1433693735, 1433694028, 1433694373, 1433694642,
1433694931, 1433695234, 1433695542, 1433695831, 1433696148,
1433696448, 1433710908, 1433714437, 1433714737, 1433715036,
1433715366, 1433715636, 1433715969, 1433716234, 1433716536,
1433716827, 1433717137, 1433717435, 1433717733, 1433718048,
1433718336, 1433718636, 1433718973, 1433719272, 1433719530,
1433719837, 1433720136, 1433720431, 1433720736, 1433721031,
1433721336, 1433721640, 1433721946, 1433722236, 1433722528,
1433722842, 1433723137, 1433723434, 1433723730, 1433724035,
1433724370, 1433724634, 1433724936, 1433725236), class = c("POSIXct",
"POSIXt"), tzone = ""), UTM_X = c(636979.2503, 636977.6583,
637402.4471, 637400.3063, 637402.3105, 637407.1977, 637406.3305,
637408.2991, 637407.1907, 637407.8414, 637406.7617, 637407.1614,
637409.8019, 637431.5235, 637465.9644, 637495.9583, 637525.2219,
637573.6033, 637645.3501, 637683.3844, 637691.6229, 637693.4815,
637693.4973, 637691.2483, 637691.9061, 637693.6377, 637692.1106,
637692.3169, 637690.9989, 637691.4503, 637693.6252, 637692.4915,
637694.9434, 637692.6685, 637692.8116, 637694.6787, 637694.4404,
637695.9109, 637696.8945, 637695.2403, 637695.4283, 637694.6085,
637693.4962, 637695.6229, 637734.7283, 637773.2897, 637774.9891,
637787.6573, 637792.285, 637807.0486, 637834.6231, 637497.3348,
637149.9982, 637145.0345, 637178.159, 637181.8251, 637181.1075,
637178.023, 637175.327, 637179.9138, 637180.2833, 637181.5512,
637185.8749, 637181.0011, 637177.401, 637177.4498, 637176.787,
637176.0093, 637175.5126, 637177.9578, 637178.5819, 637188.3911,
637188.7303, 637189.496, 637204.3885, 637195.2063, 637204.9823,
637201.5235, 637212.3355, 637274.4294, 637293.0009, 637296.3954,
637331.3382, 637358.4369, 637365.1677, 637357.5562, 637355.3896,
637345.4827, 637339.1054, 628920.3789, 628869.9781, 630028.6781,
630156.4557, 629878.756, 629658.9786, 629412.6432, 629257.5965,
629405.8967, 629113.4479, 628955.5124, 628852.0231, 628711.9202,
628632.7134, 628621.7724, 628622.2565, 628683.6018, 628771.1182,
628790.8437, 628867.7592, 628881.9794, 628830.9898, 628681.9202,
628575.3395, 628578.1836, 628656.4902, 628659.2271, 628656.689,
628660.4677, 628657.294, 628657.077, 628689.6585, 628727.0131,
628716.6979, 628703.8397, 628678.6953, 628679.3594, 628681.3549,
628625.6275, 628563.1372, 628488.425, 628482.5023, 628469.2209,
628417.9697, 628407.7352, 628405.374, 628393.143, 628394.0092,
628396.2344, 628395.05, 628395.7787, 627684.7989, 627704.889,
627702.5528, 627702.0422, 627708.7906, 627706.9374, 627687.0371,
627622.0573, 627605.7932, 627603.5707, 627587.8803, 627606.0471,
627603.2967, 627602.954, 627603.5844, 627604.1232, 627601.697,
627581.6104, 627599.7062, 627616.327, 627661.7402, 627889.446,
627883.5896, 627803.1167, 627792.5918, 627716.0886, 627720.8854,
627671.8217, 627666.9994, 627586.7035, 627584.4273, 627532.492,
627502.6326, 627430.6781, 627408.8845, 627357.5049, 627406.0466,
627427.1382, 636666.3215, 636629.7032, 637179.9041, 637187.7067,
637183.5281, 637193.2082, 637227.2331, 637290.2543, 637347.9311,
637373.0887, 637368.8923, 637371.0722, 637383.95, 637480.1799,
637510.543, 637558.428, 637676.2714, 637682.3564, 637680.8591,
637682.8516, 637680.8317, 637680.8341, 637681.9818, 637681.2897,
637681.3658, 637681.9234, 637681.8824, 637682.0629, 637684.8756,
637681.602, 637682.7548, 637680.8578, 637682.9887, 637680.2496,
637681.4629, 637682.3731, 637682.2223, 637684.1076, 637681.7127,
637681.1249, 637681.6758, 637681.595, 637682.5253, 637702.3094,
637728.9487, 637784.0853, 637776.5727, 637785.2538, 637786.6413,
637807.9935, 637834.8672, 637485.5191, 637148.5674, 637139.2974,
637174.9104, 637191.9371, 637179.4262, 637175.7715, 637176.3455,
637174.5459, 637174.2012, 637173.7462, 637177.3967, 637176.6907,
637177.8458, 637178.0774, 637178.4151, 637178.3272, 637178.2442,
637177.6655, 637176.734, 637186.2713, 637185.0998, 637197.4201,
637197.9147, 637204.1485, 637203.1784, 637204.4993, 637205.3515,
637279.9058, 637303.773, 637303.5724, 637330.3473, 637354.416,
637366.5627, 637340.7274, 637357.5505, 637350.709, 637349.689
), UTM_Y = c(3365828.581, 3365826.066, 3364992.673, 3364991.006,
3364989.036, 3364990.816, 3364989.486, 3364991.849, 3364991.37,
3364990.059, 3364989.58, 3364991.403, 3364991.536, 3364985.614,
3365030.733, 3365054.446, 3365091.064, 3365138.444, 3365289.033,
3365390.111, 3365398.839, 3365387.124, 3365390.427, 3365386.696,
3365387.104, 3365379.344, 3365386.131, 3365388.805, 3365385.152,
3365385.158, 3365386.394, 3365385.637, 3365385.48, 3365386.071,
3365385.397, 3365387.416, 3365387.269, 3365389.505, 3365389.971,
3365387.833, 3365389.676, 3365390.685, 3365385.96, 3365384.934,
3365352.152, 3365369.878, 3365376.795, 3365390.013, 3365382.689,
3365382.189, 3365410.939, 3365683.847, 3365620.829, 3365574.121,
3365527.084, 3365501.513, 3365502.801, 3365512.739, 3365514.733,
3365512.885, 3365511.016, 3365512.562, 3365510.255, 3365511.235,
3365509.494, 3365509.439, 3365509.431, 3365509.388, 3365510.678,
3365509.534, 3365511.083, 3365511.85, 3365514.659, 3365513.371,
3365525.476, 3365526.036, 3365529.429, 3365528.676, 3365513.172,
3365507.793, 3365514.623, 3365512.105, 3365504.477, 3365512.401,
3365495.238, 3365490.863, 3365441.075, 3365411.542, 3365403.003,
3371496.516, 3371594.382, 3370587.966, 3370380.241, 3370270.012,
3370346.817, 3370433.295, 3370488.189, 3370225.222, 3370122.896,
3370174.202, 3370232.298, 3370192.371, 3370255.722, 3370283.548,
3370283.21, 3370305.674, 3370344.002, 3370354.4, 3370348.973,
3370200.353, 3370078.071, 3370123.589, 3370194.686, 3370393.878,
3370500.265, 3370498.635, 3370498.882, 3370497.663, 3370499.687,
3370500.172, 3370633.763, 3370704.904, 3370839.426, 3370879.943,
3370950.842, 3370957.988, 3370963, 3371031.496, 3371082.487,
3371109.89, 3371112.17, 3371118.807, 3371167.072, 3371168.581,
3371170.127, 3371178.074, 3371177.097, 3371178.11, 3371176.777,
3371178.482, 3371566.662, 3371622.632, 3371621.252, 3371619.772,
3371623.975, 3371627.245, 3371636.71, 3371612.734, 3371598.776,
3371590.192, 3371636.009, 3371656.352, 3371656.719, 3371656.471,
3371656.755, 3371659.1, 3371656.401, 3371688.243, 3371717.065,
3371741.492, 3371755.505, 3371618.156, 3371595.308, 3371615.82,
3371560.55, 3371552.166, 3371572.884, 3371547.544, 3371530.616,
3371559.755, 3371591.63, 3371612.877, 3371657.663, 3371727.149,
3371739.263, 3371823.645, 3371912.149, 3371969.549, 3366104.602,
3365712.344, 3365494.627, 3365496.045, 3365484.575, 3365475.02,
3365485.304, 3365467.377, 3365477.805, 3365507.809, 3365510.682,
3365519.888, 3365527.19, 3365491.394, 3365490.37, 3365468.274,
3365393.413, 3365389.355, 3365386.964, 3365391.977, 3365389.125,
3365388.937, 3365389.35, 3365389.375, 3365387.159, 3365387.133,
3365386.578, 3365386.735, 3365386.161, 3365387.472, 3365387.487,
3365387.064, 3365387.977, 3365385.016, 3365387.836, 3365388.036,
3365387.048, 3365389.909, 3365387.074, 3365384.939, 3365387.717,
3365388.026, 3365388.16, 3365385.728, 3365344.996, 3365374.693,
3365377.679, 3365387.866, 3365389.823, 3365391.779, 3365410.631,
3365698.174, 3365622.297, 3365571.954, 3365511.957, 3365510.265,
3365505.086, 3365509.196, 3365512.44, 3365513.438, 3365508.777,
3365509.049, 3365509.838, 3365506.403, 3365507.748, 3365510.711,
3365509.075, 3365507.666, 3365508.152, 3365505.285, 3365498.401,
3365508.531, 3365508.483, 3365513.538, 3365520.783, 3365519.376,
3365523.92, 3365529.634, 3365529.866, 3365498.661, 3365512.941,
3365509.801, 3365503.056, 3365513.548, 3365502.683, 3365482.215,
3365438.852, 3365412.317, 3365406.363)), .Names = c("CollarID",
"DateTime", "UTM_X", "UTM_Y"), row.names = c(NA, -267L), class = "data.frame")
Commenting out incorrect transformation
# chupacabra$DateTime <-as.POSIXct(strptime(chupacabra$DateTime, format='%m/%d/%Y %H:%M:%S'),origin='1970-01-01')
chupacabra2<-as.ltraj(chupacabra[, c("UTM_X","UTM_Y")], date=chupacabra$DateTime,id=chupacabra$CollarID, typeII=TRUE)
monster1<-chupacabra2[1] #extract the first chupacabra
monster2<-chupacabra2[2] #extract the second chupacabra
proxdf <-Prox(monster1,monster2, tc=0.5*60,dc=210, local =TRUE)
Here's a sample of chupcabras that we've been tracking. We'd like to examine how often they interact with each other. This dataset has 3 individuals (but we have many many more chupcabras) and it is inefficient to pull out the animals/creatures 1 by 1 to calculate proximity. I'd like to do a for loop (for i in unique ID, perhaps) but I don't understand how to do this when the data is in ltraj format. Any assistance would be appreciated.
After removing the harmful code that incorrectly reformatted the DateTime value, I plotted the "trajectory" for the 3 2-animal combinations. Clearly animal1 did not interact with animal 2 since their ranges were disjoint. Animal 1 and animal 3 do appear to interact, since about halfway through their joint sojourns they have the roughly same trajectory:
plot( chupacabra2[[1]]$x, chupacabra2[[1]]$y, type="l",
xlim=range( c(chupacabra2[[1]]$x, chupacabra2[[3]]$x)),
ylim= range(c(chupacabra2[[1]]$y, chupacabra2[[3]]$y)))
lines( chupacabra2[[3]]$x, chupacabra2[[3]]$y, col="red")
This appears in proxdf13 as:
> monster1<-chupacabra2[1]
>
> monster3<-chupacabra2[3]
>
> proxdf13 <-Prox(monster1,monster3, tc=0.5*60,dc=210, local =TRUE)
> proxdf13
date prox
1 2015-06-07 06:01:00 549.074863
2 2015-06-07 07:20:00 104.169909
3 2015-06-07 08:10:00 6.205875
4 2015-06-07 09:05:00 14.409016
5 2015-06-07 09:20:00 11.189309
6 2015-06-07 15:40:00 6.481131
7 2015-06-07 16:25:00 4.259042
8 2015-06-07 17:15:00 10.648210
9 2015-06-07 17:35:00 4.181297
10 2015-06-07 17:41:00 7.574566
So a guess that "interact" (might) means something along the lines of "is within 15 distance units for more than 2 days in succession". So a natural function to consider would be rle:
> rle( proxdf13$prox < 15 )
Run Length Encoding
lengths: int [1:2] 2 8
values : logi [1:2] FALSE TRUE
> RL13 <- rle( proxdf13$prox < 15 )
> max( RL13$lengths [ RL13$values] )
[1] 8
And test whether this is greater than some value, say 2?
So that was the method to handle a single 2way animal-animal combination. The comments of the questioner below suggest I may have lost him in the material that follows. To get the 2way combinations of a sequence use combn:
> combn(1:3, 2)
[,1] [,2] [,3]
[1,] 1 1 2
[2,] 2 3 3
That is a matrix that will be use to generate the index values for pulling two single animal "trajectory" dataframes at a time from the 'ltraj' object using each column separately. When the apply function is used it can "loop" over etiehr rows or columns of a matrix and using 2 as the index value to apply specifieds the columns.
So putting this all together (using apply to loop over the column indices of combn-result to get the two-way combinations):
( max.days.prox <- apply( combn( seq(length(chupacabra2)), 2 ), 2,
# loops over columns of the "combinations matrix"
function(x) {
proxcomb <- Prox( chupacabra2[ x[1] ], chupacabra2[ x[2] ],
tc=0.5*60,dc=210, local =TRUE)
RLcomb <- rle( proxcomb$prox < 15 )
Interact.days <- max( RLcomb$lengths [ RLcomb$values] ) } ) )
# [1] -Inf 8 -Inf
We rbind that to the result to look at the items of interest:
> rbind(combn( seq(length(chupacabra2)), 2 ) , max.days.prox)
[,1] [,2] [,3]
1 1 2
2 3 3
max.days.prox -Inf 8 -Inf
So only the pairing of animal1 and animal3 provided evidence of an interaction. This would generalize to larger instances of ltraj-objects
Related
Why am I getting the same result from Anova and aov_car in R, and why is it different from SPSS?
I am conducting a reanalysis of some data. The dv is continuous (beta value ie neural activity) and the iv is categorical (position) with three levels (1, 2, 3). Position is set as a factor. It is repeated measures, and there are 126 observations. The original analysis was done in SPSS, and I am trying to replicate those results with R. I don't understand how to make a MRE of this, so my data from dput is at the bottom. My ANOVA results are different from those reported in the original paper (the data is identical). Specifically, they reported F(2,82) = 18.262, p = 0.00, but my table (below) is totally different. I used the Anova function, and now get the impression that I should be using aov_car but the output is the same between the two. > Anova(lm(Beta ~ Position, data = stack_ex)) Anova Table (Type II tests) Response: Beta Sum Sq Df F value Pr(>F) Position 60.57 2 1.5213 0.2225 Residuals 2448.70 123 > aov_car(Beta ~ Position + Error(Beta), data = stack_ex) Contrasts set to contr.sum for the following variables: Position Anova Table (Type 3 tests) Response: Beta Effect df MSE F ges p.value 1 Position 2, 123 19.91 1.52 .024 .222 --- Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘+’ 0.1 ‘ ’ 1 I didn't know what to use as the error term, so I put Beta. Is this the issue? Here is the data, apologies for not being able to reduce it down. > dput(stack_ex) structure(list(Beta = c(9.97627322506813, 4.51007015616003, 12.5899137493145, 5.16107528902195, 0.69934803628816, 3.05576441003722, 9.73415586595716, 3.48253752326239, 8.72271400892749, 6.58254223482513, 9.64252595049282, 6.2575247824253, 5.55086088416984, 3.26575073266956, -0.189486641765607, -6.34627220217585, 3.03699535774724, 5.38452950644857, 7.2247809046584, 1.05684383099248, 0.745997758871227, 13.4708766693015, 6.22313273382721, 7.60691743953363, 7.95869706610072, 0.0733745510036445, 5.74455260852637, 9.10243217750976, 3.83463985621549, 6.51540068169028, 6.74657874951813, 9.06748922888841, 4.18661204617864, 8.13865720827057, 4.97289378228525, 4.79399790512039, -12.5433736154914, 3.22520674616528, 4.83924807559523, 6.89780284608954, 2.01175994751707, 1.58936731656692, 8.65646845487533, 2.03332866864119, 6.59573013233866, 4.35624613417537, 3.22584501764675, 3.01812749198894, 8.67739700219412, 5.14273744714805, 7.54959191256081, 7.83244934217214, 8.67126128885367, 3.99955715822518, 2.95804569815409, 2.25327292671231, 0.258342636171449, -6.87648408967595, 1.9848049507549, 2.45033479610578, 7.41525416520838, 1.11896377050173, 0.0698315480648937, 9.90975895502056, 5.03717210651178, 4.67127493715398, 7.90306051043896, 3.0618932143297, 5.43781266582611, 8.9383987897543, 4.7982992164727, 6.90576740201611, 4.43862196057089, 9.06484925843098, 3.35645527138813, 5.42103905597134, 2.32859166774007, 3.65962841104834, -11.716124636774, 7.15256990819002, 4.02640955184303, 7.10747478179406, 2.81026958853589, 1.21494403713035, 9.06256308202033, 2.40170878761068, 6.45729748790901, 4.88232212084591, 1.55722661655526, 3.09556060018938, 6.6629967466337, 4.38848062553557, 4.38871083406173, 6.40367918458127, 6.361735558817, 4.21279189431753, 2.08838813524482, 2.21632202746396, -0.491401226521853, -7.3685373528786, 2.12839354041543, 4.22958686769682, 4.25606944426722, 0.330400668298046, 1.02776552933976, 10.6734745608271, 3.01238218831987, 4.03318609054561, 6.45849154079659, 0.45593329021199, 5.76390726591623, 7.21202360734704, 4.62140561321984, 3.72714943200746, 5.49911004676976, 9.15658405382221, 3.25231083403689, 3.67627240704932, 3.48390458422993, 2.98674297337782, -19.5189775914798, 2.59812967326379, 2.78334604762499, 3.70635047793331, -0.223282095324164, 2.17552096286021), Position = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L), .Label = c("1", "2", "3"), class = "factor")), row.names = c(NA, -126L), class = c("tbl_df", "tbl", "data.frame"))
How to separate compact letter display (CLD) in multcomp by group without changing the p-value adjustment method?
Problem I would like to plot estimated marginal means from a three-way factorial experiment with letters indicating significantly different means, adjusted for multiple comparisons. My current workflow is to fit the model with lmer(), calculate estimated marginal means with emmeans(), then implement the compact letter display algorithm with cld(). My problem is that the graph is too busy when you plot all three-way interactions on the same plot. So I would like to split up the plot and generate different sets of letters for each subplot, starting with "a". The problem is that when I use the by argument in cld to split it up, it does a separate correction for multiple comparisons within each by group. Because there are now fewer tests within each group, this results in a less conservative correction. But if I try to manually split up the output of cld() without a by group, I would have to manually re-implement the letter algorithm for each subplot. I guess I could do that but it seems cumbersome. I am trying to share this code with a client for him to modify later, so that solution would probably be too complex. Does anyone have an easy way to either: Get the output of cld() to use one combined correction for all by groups. Using a relatively simple method, reduce the compact letter display for each subgroup to the minimal necessary number of letters. Reproducible example Load packages and data. library(lme4) library(emmeans) library(multcomp) dat <- structure(list(y = c(2933.928571, 930.3571429, 210.7142857, 255.3571429, 2112.5, 1835.714286, 1358.928571, 1560.714286, 9192.857143, 3519.642857, 2771.428571, 7433.928571, 4444.642857, 3025, 3225, 2103.571429, 3876.785714, 925, 1714.285714, 3225, 1783.928571, 2223.214286, 2537.5, 2251.785714, 7326.785714, 5130.357143, 2539.285714, 6116.071429, 5808.928571, 3341.071429, 2212.5, 7562.5, 3907.142857, 3241.071429, 1294.642857, 4325, 4487.5, 2551.785714, 5648.214286, 3198.214286, 1075, 335.7142857, 394.6428571, 1605.357143, 658.9285714, 805.3571429, 1580.357143, 1575, 2037.5, 1721.428571, 1014.285714, 2994.642857, 2116.071429, 800, 2925, 3955.357143, 9075, 3917.857143, 2666.071429, 6141.071429, 3925, 1626.785714, 2864.285714, 7271.428571, 3432.142857, 1826.785714, 514.2857143, 1319.642857, 1782.142857, 2637.5, 1355.357143, 3328.571429, 1914.285714, 817.8571429, 1896.428571, 2121.428571, 521.4285714, 360.7142857, 1114.285714, 1139.285714, 7042.857143, 2371.428571, 2287.5, 4967.857143, 2180.357143, 1944.642857, 2408.928571, 5289.285714, 7028.571429, 3080.357143, 5394.642857, 5973.214286, 7323.214286, 1419.642857, 1455.357143, 4657.142857, 7069.642857, 2451.785714, 4319.642857, 5562.5, 3953.571429, 1182.142857, 1957.142857, 3796.428571, 1773.214286, 400, 871.4285714, 842.8571429, 657.1428571, 1360.714286, 1853.571429, 1826.785714, 3405.357143, 2605.357143, 5983.928571, 4935.714286, 4105.357143, 7666.071429, 3619.642857, 5085.714286, 1592.857143, 1751.785714, 5992.857143, 2987.5, 794.6428571, 3187.5, 825, 3244.642857), f1 = structure(c(4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L), .Label = c("A", "B", "C", "D"), class = "factor"), f2 = structure(c(2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L), .Label = c("foo", "bar"), class = "factor"), f3 = structure(c(4L, 3L, 2L, 1L, 3L, 4L, 1L, 2L, 4L, 2L, 1L, 3L, 3L, 2L, 4L, 1L, 3L, 1L, 4L, 2L, 2L, 4L, 3L, 1L, 2L, 4L, 1L, 3L, 2L, 3L, 1L, 4L, 3L, 4L, 1L, 2L, 3L, 2L, 4L, 1L, 2L, 1L, 3L, 4L, 1L, 2L, 4L, 3L, 2L, 1L, 3L, 4L, 3L, 1L, 4L, 2L, 4L, 2L, 3L, 1L, 1L, 3L, 2L, 4L, 3L, 4L, 1L, 2L, 1L, 4L, 3L, 2L, 3L, 1L, 4L, 2L, 1L, 3L, 4L, 2L, 4L, 3L, 1L, 2L, 1L, 3L, 4L, 2L, 3L, 1L, 4L, 2L, 4L, 1L, 3L, 2L, 2L, 3L, 4L, 1L, 4L, 1L, 2L, 3L, 4L, 1L, 3L, 2L, 1L, 2L, 4L, 3L, 1L, 2L, 4L, 3L, 1L, 4L, 2L, 3L, 1L, 3L, 4L, 2L, 1L, 3L, 2L, 4L), .Label = c("L1", "L2", "L3", "L4"), class = "factor"), block = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L), .Label = c("1", "2", "3", "4"), class = "factor")), row.names = c(NA, -128L), class = "data.frame") Fit model and get estimated marginal means. fit <- lmer(log10(y) ~ f1 * f2 * f3 + (1 | block), data = dat) emm <- emmeans(fit, ~ f1 + f2 + f3, mode = 'Kenward-Roger', type = 'response') Version 1 In this version, I take the CLD as a whole which correctly uses the Sidak adjustment for 496 tests. However let's say I wanted to plot only those rows where f2 == 'bar'. The letters are no longer correct because some are redundant (less than 8 are needed). Is there any function that can reduce the letters down? cldisplay1 <- cld(emm, adjust = 'sidak', Letters = letters) subset(as.data.frame(cldisplay1), f2 == 'bar') # correct comparisons but contains redundant letters output f1 f2 f3 response SE df lower.CL upper.CL .group 8 D bar L1 365.6732 76.1231 96 185.9699 719.0244 a 24 D bar L3 582.8573 121.3349 96 296.4229 1146.0742 ab 16 D bar L2 682.9238 142.1659 96 347.3136 1342.8353 ab 7 C bar L1 898.1560 186.9714 96 456.7740 1766.0470 abcd 6 B bar L1 1627.7069 338.8438 96 827.8006 3200.5652 bcdefg 15 C bar L2 1635.4393 340.4534 96 831.7330 3215.7694 bcdefg 32 D bar L4 1746.6052 363.5951 96 888.2685 3434.3552 bcdefg 31 C bar L4 2348.6629 488.9270 96 1194.4562 4618.1832 cdefgh 21 A bar L3 2499.6772 520.3640 96 1271.2573 4915.1230 cdefgh 5 A bar L1 2545.4594 529.8946 96 1294.5407 5005.1448 cdefgh 23 C bar L3 2561.0138 533.1326 96 1302.4512 5035.7294 cdefgh 30 B bar L4 3158.6969 657.5538 96 1606.4140 6210.9556 efgh 22 B bar L3 3364.9438 700.4887 96 1711.3047 6616.4994 efgh 14 B bar L2 3411.4009 710.1598 96 1734.9313 6707.8482 efgh 13 A bar L2 3769.4223 784.6900 96 1917.0098 7411.8269 efgh 29 A bar L4 7006.3740 1458.5342 96 3563.2217 13776.6551 h Version 2 In this version, I use the by argument to cld() to split by f2. This reduces the letters within each group, but the Sidak adjustment is now less conservative. For example, row 8 and row 16 are not significantly different at the adjusted alpha-level from the comparison above, but now they are different. But I do not want to change the tests used, just to plot only a subset of the data. Is there a way to specify the number of tests I'm adjusting for as a whole, even though cld is split up with by groups? cldisplay2 <- cld(emm, adjust = 'sidak', by = 'f2', Letters = letters) subset(as.data.frame(cldisplay2), f2 == 'bar') output f1 f2 f3 response SE df lower.CL upper.CL .group 8 D bar L1 365.6732 76.1231 96 185.9699 719.0244 a 24 D bar L3 582.8573 121.3349 96 296.4229 1146.0742 ab 16 D bar L2 682.9238 142.1659 96 347.3136 1342.8353 abc 7 C bar L1 898.1560 186.9714 96 456.7740 1766.0470 abcd 6 B bar L1 1627.7069 338.8438 96 827.8006 3200.5652 bcde 15 C bar L2 1635.4393 340.4534 96 831.7330 3215.7694 bcde 32 D bar L4 1746.6052 363.5951 96 888.2685 3434.3552 cde 31 C bar L4 2348.6629 488.9270 96 1194.4562 4618.1832 de 21 A bar L3 2499.6772 520.3640 96 1271.2573 4915.1230 def 5 A bar L1 2545.4594 529.8946 96 1294.5407 5005.1448 def 23 C bar L3 2561.0138 533.1326 96 1302.4512 5035.7294 def 30 B bar L4 3158.6969 657.5538 96 1606.4140 6210.9556 ef 22 B bar L3 3364.9438 700.4887 96 1711.3047 6616.4994 ef 14 B bar L2 3411.4009 710.1598 96 1734.9313 6707.8482 ef 13 A bar L2 3769.4223 784.6900 96 1917.0098 7411.8269 ef 29 A bar L4 7006.3740 1458.5342 96 3563.2217 13776.6551 f
With the two separate tables (or plots?) you are displaying a total of 90 + 90 = 180 comparisons. If you want an overall multiplicity adjustment for all of these 180 comparisons, you need to be considerably less conservative than for 496 comparisons. However, it is possible to speccify a different value of level so that the Sidak adjustment works out correctly. For example, if you want the overall alpha to be 0.05, use cld(emm, adjust = 'sidak', by = 'f2', Letters = letters, alpha = 1 - sqrt(0.95)) With this, you are specifying alpha = 0.02532. Note that if p.adj = 1 - (1 - p)^90 < 1 - sqrt(.95) then (1 - p)^90 > sqrt(.95) so that (1 - p)^180 > .95 thus 1 - (1 - p)^180 < .05 That is, by splitting the CLD table into two parts showing 90 comparisons each, we correctly apply the Sidak adjustment to correct for the 180 comparisons total at a significance level of .05. Enhancement Another idea based on this that results in a less conservative adjustment is to specify the Tukey adjustment instead: cld(emm, adjust = 'tukey', by = 'f2', Letters = letters, alpha = 1 - sqrt(0.95)) Thus, each separate table has an exact familywise error rate of 1 - sqrt(0.05); and we used the Sidak adjustment (slightly conservative) so that the error rate for the whole family of 180 tests is less than 0.05.
Setting up an Mlogit in R with many observations for each category
I'm trying to use Mlogit in R, I'm a little new to logits, and I'm having trouble setting up my problem in the Mlogit framework. I'm actually not entirely sure that mlogit is the right approach. Here is an analogous problem. Consider a baseball dataset, with an outcome variable that takes on "out" "single" "double" "triple" and "homerun." For explanatory variables, we have the name of the batter, the name of the pitcher, and the stadium. There are hundreds of observations for each batter, including many with the batter facing the same pitcher. I figured this is definitely a multinomial logit because I have multiple categorical outcomes, but I am not sure because all of the documentation seems to be dealing with "choices" between alternatives, which this isn't really. I tried to start my logit model by having a factor variable for the hitter, another one for the pitcher, and another one for the stadium. When I tried this in R, I get Error in row.names<-.data.frame(*tmp*, value = value) : invalid 'row.names' length With some googling I think maybe it is expecting only one observation for each combination of hitter, pitcher, and park? Maybe not? What am I doing wrong? How should I set this up? Edit: Example of data here https://docs.google.com/spreadsheets/d/19fiq_QEMj4nAPcTqIRxeaYNPgqeHxKAEuPrfHMeIJ7o/edit?usp=sharing
Here are some suggestions on how you can start analyzing your data. # Your dataset dts <- structure(list(outcome = c(1L, 1L, 2L, 3L, 1L, 3L, 2L, 3L, 3L, 3L, 3L, 1L, 2L, 2L, 2L, 1L, 3L, 2L, 2L, 2L, 1L, 2L, 3L, 2L, 2L, 2L, 2L, 1L, 1L, 2L, 3L, 2L, 3L, 1L, 2L, 2L, 3L, 2L, 3L, 3L, 3L, 2L, 1L, 1L, 1L, 2L, 3L, 2L, 1L), hitter = structure(c(3L, 3L, 3L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 3L, 3L, 2L, 2L, 2L, 1L, 1L, 1L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L), .Label = c("james", "jill", "john"), class = "factor"), pitcher = structure(c(3L, 3L, 1L, 1L, 1L, 1L, 2L, 2L, 3L, 2L, 2L, 2L, 2L, 2L, 3L, 1L, 1L, 2L, 2L, 3L, 3L, 3L, 1L, 1L, 1L, 2L, 2L, 3L, 2L, 1L, 2L, 3L, 2L, 3L, 2L, 1L, 1L, 2L, 2L, 1L, 3L, 3L, 1L, 2L, 2L, 1L, 1L, 2L, 2L ), .Label = c("bill", "bob", "brett"), class = "factor"), place = structure(c(3L, 3L, 3L, 3L, 3L, 3L, 3L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 5L, 5L, 5L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L ), .Label = c("ca", "co", "dc", "ny", "tn"), class = "factor")), .Names = c("outcome", "hitter", "pitcher", "place"), class = "data.frame", row.names = c(NA, -49L)) # Estimation of a multinomial logistic regression model library(mlogit) dts.wide <- mlogit.data(dts, choice="outcome", shape="wide") fit.mlogit <- mlogit(outcome ~ 1 | hitter+pitcher+place, data=dts.wide) # Results library(stargazer) stargazer(fit.mlogit, type="text") # Model coefficients with standard errors and statistical significance (stars) ========================================== Dependent variable: --------------------------- outcome ------------------------------------------ 2:(intercept) 19.456 (3,056.626) 3:(intercept) 35.179 (4,172.540) 2:hitterjill -17.543 (3,056.625) 3:hitterjill -33.117 (4,172.540) 2:hitterjohn -0.188 (0.996) 3:hitterjohn -1.410 (1.056) 2:pitcherbob -0.070 (1.005) 3:pitcherbob -1.270 (1.091) 2:pitcherbrett -0.908 (1.063) 3:pitcherbrett -2.284* (1.257) 2:placeco -1.655 (1.557) 3:placeco -17.688 (2,840.270) 2:placedc -19.428 (3,056.626) 3:placedc -34.479 (4,172.540) 2:placeny -18.802 (3,056.625) 3:placeny -32.873 (4,172.540) 2:placetn -18.885 (3,056.626) 3:placetn -32.140 (4,172.540) ------------------------------------------ Observations 49 R2 0.155 Log Likelihood -44.605 LR Test 16.388 (df = 18) ========================================== Note: *p<0.1; **p<0.05; ***p<0.01 More details on the estimation of multinomial logistic models in R are available here.
split block in R lmer
A factorial combination of 16 treatments (4*2*2) was replicated three times and laid out in a strip-split block. Treatments consisted of eight site preparations (4*2) applied as whole plot treatments and two levels of weeding(weeding/no-weeding) were applied randomly to subplots. The analysis was run in Genstat giving the following results: Variate: result Source of variation d.f. s.s. m.s. v.r. F pr. Rep stratum 2 35.735 17.868 Rep.Burning stratum Burning 1 0.003 0.003 0.00 0.972 Residual 2 3.933 1.966 1.53 Rep.Site_prep stratum Site_prep 3 7.981 2.660 0.45 0.727 Residual 6 35.477 5.913 4.61 Rep.Burning.Site_prep stratum Burning.Site_prep 3 2.395 0.798 0.62 0.626 Residual 6 7.691 1.282 0.60 Rep.Burning.Site_prep.*Units* stratum Weeding 1 13.113 13.113 6.13 0.025 Burning.Weeding 1 0.486 0.486 0.23 0.640 Site_prep.Weeding 3 17.703 5.901 2.76 0.076 Burning.Site_prep.Weed.3 3.425 1.142 0.53 0.666 Residual 16 34.248 2.141 Total 47 162.190 I want to repeat these results in R. I used both the base::aov function and the lmerTest::lmer function. I managed to get the correct results with aov using function result ~ Burning * Weeding * Site.prep + Error(Rep/Burning*Site.prep). With lmer I used the function result ~ Burning*Site.prep*Weeding+(1|Rep/(Burning:Site.prep)) giving me only partially correct results. The SS values and the F-values for Burning, Site.prep and Burning:Site.prep deviated (although not too much)from the Genstat results, but the Weeding and Weeding interactions gave the same SS and F-valus as the Genstat output. I would like to know how I should specify the lmer model to reproduce the Genstat and aov results. Data and code below: x <- structure(list( Rep = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 3L, 3L, 3L, 3L, 2L, 2L, 2L, 2L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L), .Label = c("1", "2", "3" ), class = "factor"),Burning = structure(c(1L, 1L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 1L, 1L), .Label = c("Burn", "No-burn"), class = "factor"), Site.prep = structure(c(4L, 4L,4L, 4L, 2L, 2L, 2L, 2L, 3L, 3L, 3L, 3L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 4L, 4L, 4L, 4L, 2L, 2L, 2L, 2L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 1L, 1L, 1L, 1L, 4L, 4L, 4L, 4L), .Label = c("Chop_Pit", "Chop_Rip", "Pit", "Rip"), class = "factor"), Weeding = structure(c(1L, 2L, 1L, 2L, 1L, 2L, 2L, 1L, 2L, 1L, 2L, 1L, 1L, 2L, 2L, 1L, 1L, 2L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 1L, 2L, 2L, 1L, 1L, 2L, 1L, 2L, 2L, 1L, 1L, 2L, 2L, 1L, 1L, 2L, 1L, 2L, 1L, 2L), .Label = c("Weedfree", "Weedy"), class = "factor"), Dbh14 = c(27.4, 28.4083333333333, 27.7066666666667, 27.3461538461538, 28.6, 28.3333333333333, 27.0909090909091, 27.8076923076923, 27.1833333333333, 27.5461538461538, 24.3076923076923, 29.3461538461538, 27.4, 25.1, 26.61, 28.0461538461538, 27.71, 25.2533333333333, 25.3833333333333, 24.2307692307692, 24.2533333333333, 24.95, 24.34375, 26.9909090909091, 24.775, 25.9076923076923, 25.1666666666667, 25.9933333333333, 27.0466666666667, 30.5625, 27.36, 25.2636363636364, 29.6846153846154, 27.7, 28.3071428571429, 29.4857142857143, 27.025, 30.1, 31.2454545454545, 24.2888888888889, 28.4875, 29.23, 30, 28.5, 29.3615384615385, 27.45, 28.8153846153846, 29.1866666666667)), .Names = c("Rep", "Burning", "Site.prep", "Weeding", "result"), class = "data.frame", row.names = c(NA, -48L)) model1 <- aov(result ~ Burning* Weeding*Site.prep+ Error(Rep/Burning*Site.prep), data=x) summary(model1) model2 <- lmer(result ~ Burning*Site.prep*Weeding+(1|Rep/(Burning:Site.prep)),data=x) anova(model2)
Applying the three-way split-plot-factorial ANOVA example from the site mentioned by #cuttlefish44, leads to: library(lme4) library(nlme) m1 <- aov(result ~ Weeding*Burning*Site.prep + Error(Rep/Burning*Site.prep), data=x) m2 <- lmer(result ~ Weeding*Burning*Site.prep + (1|Rep) + (1|Burning:Rep) + (1|Site.prep:Rep), data=x) m3 <- anova(lme(result ~ Weeding*Burning*Site.prep, random=list(Rep=pdBlocked(list(~1, pdIdent(~Burning-1), pdIdent(~Site.prep-1)))), method="ML", data=x)) summary(m1) anova(m2) m3 Except for Site.prep, the results match. Moreover, the results between lmer() and lme() are pretty similar (also for Site.prep). I'm not sure whether this is the result of differences in modelling approaches: the multi-level approach takes both within and between effects into account.
Different behavior for edit(data)/vi(data) in RStudio on OS X
I have RStudio on an OS X box, and have just reinstalled X11 to fix some of the reported problems with Yosemite breaking symlinks. I go forth to the prompt and enter: > edit(iris[[5]]) and yield a new window with a wonderful structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L), .Label = c("setosa", "versicolor", "virginica"), class = "factor") and, refreshingly, I get the same result with > vi(iris[[5]]) However, moving on to data <- read.table(text="160525.000,38.63 160526.086,38.63 160526.604,38.62 160526.710,38.63 160526.343,38.62", sep=",") date = "20150824" data$Dates <- ymd_hms(paste(date,data$V1)) and attempting > edit(data) I am returned > edit(data) Error in edit.data.frame(data) : can only handle vector and factor elements yet > vi(data) structure(list(V1 = c(160525, 160526.086, 160526.604, 160526.71, 160526.343), V2 = c(38.63, 38.63, 38.62, 38.63, 38.62), Dates = structure(c(1440432325, 1440432326.086, 1440432326.604, 1440432326.71, 1440432326.343 ), tzone = "UTC", class = c("POSIXct", "POSIXt"))), .Names = c("V1", "V2", "Dates"), row.names = c(NA, -5L), class = "data.frame") in its own, separate, happy window. Why would edit and vi, in RStudio, both invoking the same document editor - On the OS X GUI the argument [Editor] is ignored and the document editor is always used. behave differently? I am not (to the best of my knowledge) invoking vi with my call to vi, but rather the same editor as edit. Why would there be a difference in response to the input? In case anyone thinks it relevant - > R.version _ platform x86_64-apple-darwin13.4.0 arch x86_64 os darwin13.4.0 system x86_64, darwin13.4.0 status major 3 minor 2.1 year 2015 month 06 day 18 svn rev 68531 language R version.string R version 3.2.1 (2015-06-18) nickname World-Famous Astronaut