Hey Jean,
This appears to have worked!! Thank you so much, this is an immense help to
me.
Best
On Tue, Mar 26, 2013 at 11:33 AM, Adams, Jean <jvad...@usgs.gov> wrote:
> Thanks. That helps.
>
> Using the example data you sent, I came up with this loop. See if it does
> the trick for you.
>
> niter <- 2
> nspp <- 11
> ntrees <- 22
> lnbm <- matrix(NA, niter, ntrees)
> for(i in 1:niter) {
> for(j in 1:ntrees) {
> k <- j1data$its==i & j1data$spp==predictdata$spp[j]
> lnbm[i, j] <- j1data$b0[k] + j1data$b1[k]*predictdata$lndbh[j]
> }
> }
>
> Jean
>
> P.S. You should cc rhelp on all correspondence, so other folks can see
> the messages unfold.
>
>
>
> On Tue, Mar 26, 2013 at 8:41 AM, Curtis Burkhalter <
> curtisburkhal...@gmail.com> wrote:
>
>> Hey Jean,
>>
>> Sorry for not including this before, here is the first 22 rows of j1data:
>>
>> its spp b0 b1
>> 1 1 1 -1.707 2.124
>> 2 1 2 -2.679 2.478
>> 3 1 3 -1.103 1.984
>> 4 1 4 -1.895 2.258
>> 5 1 5 -1.372 2.127
>> 6 1 6 -2.486 2.372
>> 7 1 7 -2.261 2.336
>> 8 1 8 -1.855 2.147
>> 9 1 9 -1.923 2.264
>> 10 1 10 -1.772 2.250
>> 11 1 11 -2.540 2.272
>> 12 2 1 -1.862 2.163
>> 13 2 2 -2.189 2.329
>> 14 2 3 -1.253 2.029
>> 15 2 4 -2.533 2.443
>> 16 2 5 -1.262 2.085
>> 17 2 6 -2.800 2.483
>> 18 2 7 -2.547 2.426
>> 19 2 8 -1.853 2.155
>> 20 2 9 -2.026 2.297
>> 21 2 10 -1.966 2.297
>> 22 2 11 -2.850 2.345
>> >
>> and here is the first 22 rows of predictdata:
>>
>> spp dbh height bm lndbh dbh2 ln_h h2 logbm
>> logdbh
>> 1 1 16.3 9.4 66.64 2.791165 265.69 2.240710 88.36 1.823735
>> 1.212188
>> 2 1 18.3 10.9 85.75 2.906901 334.89 2.388763 118.81 1.933234
>> 1.262451
>> 3 1 19.3 7.1 61.25 2.960105 372.49 1.960095 50.41 1.787106
>> 1.285557
>> 4 1 19.6 13.0 106.82 2.975530 384.16 2.564949 169.00 2.028653
>> 1.292256
>> 5 1 21.6 13.0 121.52 3.072693 466.56 2.564949 169.00 2.084648
>> 1.334454
>> 6 1 22.9 11.7 127.40 3.131137 524.41 2.459589 136.89 2.105169
>> 1.359835
>> 7 1 24.6 13.6 150.43 3.202746 605.16 2.610070 184.96 2.177334
>> 1.390935
>> 8 1 24.9 13.8 162.19 3.214868 620.01 2.624669 190.44 2.210024
>> 1.396199
>> 9 1 25.9 19.4 269.01 3.254243 670.81 2.965273 376.36 2.429768
>> 1.413300
>> 10 1 26.2 16.8 237.65 3.265759 686.44 2.821379 282.24 2.375938
>> 1.418301
>> 11 1 26.2 19.2 299.39 3.265759 686.44 2.954910 368.64 2.476237
>> 1.418301
>> 12 1 27.7 11.3 203.84 3.321432 767.29 2.424803 127.69 2.309289
>> 1.442480
>> 13 1 27.9 15.6 255.78 3.328627 778.41 2.747271 243.36 2.407867
>> 1.445604
>> 14 1 27.9 17.7 235.20 3.328627 778.41 2.873565 313.29 2.371437
>> 1.445604
>> 15 1 28.7 18.8 279.79 3.356897 823.69 2.933857 353.44 2.446832
>> 1.457882
>> 16 1 28.7 20.2 326.83 3.356897 823.69 3.005683 408.04 2.514322
>> 1.457882
>> 17 1 29.0 14.4 240.10 3.367296 841.00 2.667228 207.36 2.380392
>> 1.462398
>> 18 1 29.2 16.9 305.27 3.374169 852.64 2.827314 285.61 2.484684
>> 1.465383
>> 19 1 30.0 12.9 211.19 3.401197 900.00 2.557227 166.41 2.324673
>> 1.477121
>> 20 1 30.7 21.8 385.14 3.424263 942.49 3.081910 475.24 2.585619
>> 1.487138
>> 21 1 31.0 15.0 282.24 3.433987 961.00 2.708050 225.00 2.450619
>> 1.491362
>> 22 1 31.2 20.0 388.57 3.440418 973.44 2.995732 400.00 2.589469
>> 1.494155
>>
>> Also here is a little further clarification of what is happening when I
>> run the code and some abbreviated ouput:
>>
>> [,1] [,2] [,3] [,4] [,5]
>> [1,] 44.76950 58.23457 65.71707 68.06092 84.87327
>> [2,] 44.76950 58.23457 65.71707 68.06092 84.87327
>> [3,] 44.76950 58.23457 65.71707 68.06092 84.87327
>> [4,] 44.76950 58.23457 65.71707 68.06092 84.87327
>> [5,] 44.76950 58.23457 65.71707 68.06092 84.87327
>> [6,] 44.76950 58.23457 65.71707 68.06092 84.87327
>> [7,] 44.76950 58.23457 65.71707 68.06092 84.87327
>> [8,] 44.76950 58.23457 65.71707 68.06092 84.87327
>> [9,] 44.76950 58.23457 65.71707 68.06092 84.87327
>> [10,] 44.76950 58.23457 65.71707 68.06092 84.87327
>> [11,] 44.76950 58.23457 65.71707 68.06092 84.87327
>> [12,] 88.23264 114.99591 129.88904 134.55690 168.07243
>> [13,] 88.23264 114.99591 129.88904 134.55690 168.07243
>> [14,] 88.23264 114.99591 129.88904 134.55690 168.07243
>> [15,] 88.23264 114.99591 129.88904 134.55690 168.07243
>> [16,] 88.23264 114.99591 129.88904 134.55690 168.07243
>> [17,] 88.23264 114.99591 129.88904 134.55690 168.07243
>> [18,] 88.23264 114.99591 129.88904 134.55690 168.07243
>> [19,] 88.23264 114.99591 129.88904 134.55690 168.07243
>> [20,] 88.23264 114.99591 129.88904 134.55690 168.07243
>> [21,] 88.23264 114.99591 129.88904 134.55690 168.07243
>> [22,] 88.23264 114.99591 129.88904 134.55690 168.07243
>>
>> So it appears that it is calculating the same value for biomass 11x using
>> each iteration of the regression coefficients and then moves on. For the
>> predicted biomass located in [1,1] I know this is correct b/c I calculated
>> it by hand just to see if the code was working correctly. Moving beyond
>> that though the values contained within [1,2:5] are incorrect for the
>> corresponding trees and the subsequent values contained in [12:22,] are
>> incorrect so I know that it's not then using the 2nd iteration of the
>> regression coefficients for each tree. So to sum that up, it seems to work
>> correctly in predicting biomass for the first tree using the first
>> iteration of the regression coefficients, but beyond that something goes
>> wrong. I hope this helps in making my point clearer.
>>
>> Thank you
>>
>>
>> On Tue, Mar 26, 2013 at 8:33 AM, Adams, Jean <jvad...@usgs.gov> wrote:
>>
>>> It would be helpful to see what your data frames look like ... perhaps
>>> you could share the first few rows of each with us.
>>>
>>> dput(head(j1data))
>>>
>>> dput(head(predictdata))
>>>
>>> Jean
>>>
>>>
>>> On Mon, Mar 25, 2013 at 9:29 PM, Curtis Burkhalter <
>>> curtisburkhal...@gmail.com> wrote:
>>>
>>>> Hello,
>>>>
>>>> I'm working on a problem using nested for-loops and I don't know if
>>>> it's a
>>>> problem with the order of the loops or something within the loop so any
>>>> help with the problem would be appreciated. To briefly set up the
>>>> problem.
>>>> I have 259 trees (from 11 different species, of unequal count for each
>>>> species) of which I am trying to predict biomass. For each tree
>>>> species I
>>>> have 10000 iterations for the regression coefficients, which were
>>>> estimated
>>>> previously in WinBUGS, and what I'm trying to do is for each tree I
>>>> want to
>>>> predict the biomass using each species-specific iteration of the
>>>> regression
>>>> coefficients, so that ultimately each tree has 10000 estimates of
>>>> biomass,
>>>> organized into a 10000x259 matrix. The input data used in the model
>>>> equation is stored in two separate files and I don't think this is
>>>> creating
>>>> problems, but I thought it might be worth mentioning. I've pasted the
>>>> code
>>>> below and if any additional info is needed please write back and I will
>>>> post it.
>>>>
>>>> #read in the model output data from Jenkins eq. 1 under "data"
>>>> #read in the prediction data set under "predict data"
>>>>
>>>> j1data=read.delim("reduced_j1_forR.txt",header=T)
>>>> predictdata=read.delim("predictset_forR.txt",header=T)
>>>> j1data=j1data[,2:4]
>>>> its=c(rep(1:10000,each=11))
>>>> j1data=cbind(its,j1data)
>>>>
>>>> #set up a matrix full of zeros "lnbm" where prediction results are
>>>> placed
>>>> #set up for loop that first loops over iteration, then species
>>>> #(total iterations=10000it/spp*11spp=110000 iterations)and then over
>>>> each
>>>> tree
>>>> #(total # of trees=259)
>>>>
>>>> niter=10000
>>>> nspp=11
>>>> ntrees=259
>>>>
>>>> lnbm=matrix(0,10000,259)
>>>> k=numeric()
>>>> for (i in 1:ntrees)
>>>> {
>>>> for (j in 1:nspp)
>>>> {
>>>> for (m in 1:niter)
>>>> {
>>>> k=((j1data$its[m]-1)*1000)+(j1data$spp[j])
>>>> #print(k)
>>>> lnbm[m,i]=j1data$b0[k]+j1data$b1[k]*predictdata$lndbh[i]
>>>> }
>>>> }
>>>> }
>>>>
>>>>
>>>>
>>>> Thanks
>>>>
>>>> [[alternative HTML version deleted]]
>>>>
>>>> ______________________________________________
>>>> R-help@r-project.org mailing list
>>>> https://stat.ethz.ch/mailman/listinfo/r-help
>>>> PLEASE do read the posting guide
>>>> http://www.R-project.org/posting-guide.html
>>>> and provide commented, minimal, self-contained, reproducible code.
>>>>
>>>
>>>
>>
>>
>> --
>> Curtis Burkhalter
>>
>
>
--
Curtis Burkhalter
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
