add alevels, tests, examples, version++

[tdhock]

> ichr <- nc::capture_melt_single(
+ iris[1,],
+ part="Sepal|Petal",
+ "[.]",
+ dim="Length|Width")
> factor(ichr$part)#default factor levels are alphabetical.
[1] Sepal Sepal Petal Petal
Levels: Petal Sepal
> (ifac <- nc::capture_melt_single(
+ iris[1,],
+ part=nc::alevels("Sepal","Petal"),
+ "[.]",
+ dim=nc::alevels("Length","Width")))
   Species   part    dim value
    <fctr> <fctr> <fctr> <num>
1:  setosa  Sepal Length   5.1
2:  setosa  Sepal  Width   3.5
3:  setosa  Petal Length   1.4
4:  setosa  Petal  Width   0.2
> ifac$part #factor with levels in same order as given in alevels().
[1] Sepal Sepal Petal Petal
Levels: Sepal Petal
> tv_wide <- data.frame(
+ id=0,
+ train.classif.logloss = 1, train.classif.ce = 2,
+ valid.classif.logloss = 3, valid.classif.ce = 4)
> nc::capture_melt_single(
+ tv_wide,
+ set=nc::alevels(valid="validation", train="subtrain"),
+ "[.]classif[.]",
+ measure=nc::alevels(ce="error_prop", auc="AUC", "logloss"))
      id        set    measure value
   <num>     <fctr>     <fctr> <num>
1:     0   subtrain    logloss     1
2:     0   subtrain error_prop     2
3:     0 validation    logloss     3
4:     0 validation error_prop     4
> nc::capture_melt_single(
+ tv_wide,
+ set_chr=list(set_fac=nc::alevels(valid="validation", train="subtrain")),
+ "[.]classif[.]",
+ measure_chr=list(measure_fac=nc::alevels(ce="error_prop", auc="AUC", "logloss")))
      id set_chr    set_fac measure_chr measure_fac value
   <num>  <char>     <fctr>      <char>      <fctr> <num>
1:     0   train   subtrain     logloss     logloss     1
2:     0   train   subtrain          ce  error_prop     2
3:     0   valid validation     logloss     logloss     3
4:     0   valid validation          ce  error_prop     4