[R] asign variables in a "for" loop

[Jim Lemon]

On 06/13/2012 11:41 PM, Nympha Nymphaea wrote:
> Dear R-helpers,
>
> I'm stuck with a little problem that surely has an easy solution but I
> can't think of a way to solve it. I'd really appreciate any help you can
> offer me!
>
> I'll provide a small example. Given a dataframe data.txt that looks like
> this:
>
> ID    freq    Var    Var_mean    Ratio_mean    Var_median
> Ratio_median    Var_sum    Ratio_min    Var_max    Ratio_max    Var_min
> Ratio_sum
> 134    5    2.140    0.447    4.784    0.272    7.881    2.237    0.957
> 0.833    2.568    0.187    11.437
> 30    4    1.743    0.450    3.873    0.358    4.869    1.799    0.968
> 0.915    1.904    0.169    10.329
> 137    4    1.401    0.304    4.614    0.310    4.514    1.215    1.153
> 0.480    2.921    0.114    12.268
> 9    3    0.065    0.023    2.849    0.021    3.108    0.069    0.950
> 0.030    2.137    0.017    3.799
> 35    3    0.192    0.067    2.849    0.025    7.756    0.202    0.950
> 0.153    1.258    0.025    7.756
> 95    3    1.365    0.360    3.792    0.335    4.073    1.080    1.264
> 0.484    2.820    0.261    5.237
> 120    3    5.171    1.891    2.735    0.542    9.532    5.672    0.912
> 4.970    1.040    0.160    32.408
> 123    3    0.721    0.226    3.182    0.155    4.661    0.679    1.061
> 0.372    1.939    0.153    4.704
> 133    3    3.242    0.918    3.531    0.921    3.519    2.754    1.177
> 1.551    2.090    0.282    11.512
> 136    3    1.058    0.371    2.850    0.337    3.141    1.113    0.950
> 0.449    2.358    0.328    3.225
> 140    3    5.935    1.877    3.162    0.327    18.167    5.630    1.054
> 5.114    1.160    0.189    31.332
> 141    3    0.974    0.325    2.997    0.160    6.092    0.975    0.999
> 0.661    1.473    0.154    6.330
> 147    3    3.207    0.798    4.018    0.951    3.373    2.394    1.339
> 1.148    2.793    0.296    10.840
> 2    2    3.859    2.020    1.911    2.020    1.911    4.039    0.955
> 3.714    1.039    0.326    11.851
> 8    2    0.017    0.009    1.900    0.009    1.900    0.018    0.950
> 0.014    1.266    0.005    3.799
> 10    2    0.096    0.060    1.603    0.060    1.603    0.120    0.802
> 0.060    1.599    0.060    1.607
> 26    2    7.308    0.504    14.500    0.504    14.500    1.008    7.250
> 0.813    8.990    0.195    37.459
> 46    2    9.070    7.542    1.203    7.542    1.203    15.085    0.601
> 7.576    1.197    7.509    1.208
> 49    2    9.485    5.035    1.884    5.035    1.884    10.070    0.942
> 9.406    1.008    0.664    14.289
> 63    2    21.308    13.956    1.527    13.956    1.527    27.912
> 0.763    14.148    1.506    13.764    1.548
>
> And I need to calculate the mean of every column whose name starts with
> Ratio_.
> That is, I need to calculate the mean of the columns Ratio_mean (3.4882),
> Ratio_median (5.2607), Ratio_sum (1.29985), Ratio_min (2.1533), Ratio_max
> (11.1469).
> But I would like to have the results just at once, in one column or one
> after the other (and not to run the code for every single result).
> I do have almost 200 columns that start with Ratio_, so doing it by hand
> it's really burdensome...
>
> I tried with a "for" loop:
>
>
> for (x in 1:5) {
>     x<- with(data, c(Ratio_mean, Ratio_median, Ratio_sum, Ratio_max,
> Ratio_min))
>     Mean<- with(data mean(x))
>     print(Mean)
>   }
>
>
> But the result is:
> [1] 4.66979
> [1] 4.66979
> [1] 4.66979
> [1] 4.66979
> [1] 4.66979
>
> which is the mean of all the columns together. I would like to obtain the
> mean of every column in each line:
>
> [1] 3.4882
> [1] 5.2607
> [1] 1.29985
> [1] 2.1533
> [1] 11.1469
>
>
> I think my problem is that I don't know how to assign correctly the
> variable x. Does anyone know if that is possible or an alternative way to
> get that result?

Hi Nympha,
Try this:

nn<-read.table("nn.dat",header=TRUE)
for(variable in grep("Ratio_",names(nn),fixed=TRUE))
  print(mean(nn[,variable]))

Jim