# [R] factor interaction boxplot ordering by median

Sergio Fonda sergio.fonda99 at gmail.com
Sat Sep 5 15:17:25 CEST 2015

```Great! Thank you
All the best
SF
Il 05/set/2015 15:08, "Marc Schwartz" <marc_schwartz at me.com> ha scritto:

>
> > On Sep 5, 2015, at 7:29 AM, Sergio Fonda <sergio.fonda99 at gmail.com>
> wrote:
> >
> > I would to visualize in boxplot a data frame with two factors ordering
> one
> > factor with the median.
> > As example,suppose to have the InsectSprays dataframe, where an
> "operator"
> > factor with two levels, op1 and op2, has been added as shown at bottom
> here.
> > How may be generated a boxplot showing boxes for the interaction
> spray*op,
> > ordered according to the operators' count median for every spray ?
> > Thanks in advance for any help!
> > Sergio
> > ________________________________
> > Modified InsectSprays dataframe:
>
> <snip>
>
> Hi,
>
> There is actually an example of reordering factor levels by a calculated
> numeric value using the InsectSprays data frame in ?boxplot using ?reorder.
> An interaction can be created by using ?interaction.
>
> Given your data above, in a data frame “DF”:
>
> DF <- structure(list(count = c(10L, 7L, 20L, 14L, 14L, 12L, 10L, 23L,
> 17L, 20L, 14L, 13L, 11L, 17L, 21L, 11L, 16L, 14L, 17L, 17L, 19L,
> 21L, 7L, 13L, 0L, 1L, 7L, 2L, 3L, 1L, 2L, 1L, 3L, 0L, 1L, 4L,
> 3L, 5L, 12L, 6L, 4L, 3L, 5L, 5L, 5L, 5L, 2L, 4L, 3L, 5L, 3L,
> 5L, 3L, 6L, 1L, 1L, 3L, 2L, 6L, 4L, 11L, 9L, 15L, 22L, 15L, 16L,
> 13L, 10L, 26L, 26L, 24L, 13L), spray = structure(c(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, 3L, 3L, 3L, 3L, 3L, 3L, 3L,
> 3L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 5L, 5L, 5L,
> 5L, 5L, 5L, 5L, 5L, 5L, 5L, 5L, 5L, 6L, 6L, 6L, 6L, 6L, 6L, 6L,
> 6L, 6L, 6L, 6L, 6L), .Label = c("A", "B", "C", "D", "E", "F"), class =
> "factor"),
>     op = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L,
>     2L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 1L,
>     1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L,
>     1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 2L,
>     2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L,
>     2L, 2L), .Label = c("op1", "op2"), class = "factor")), .Names =
> c("count",
> "spray", "op"), class = "data.frame", row.names = c("1", "2",
> "3", "4", "5", "6", "7", "8", "9", "10", "11", "12", "13", "14",
> "15", "16", "17", "18", "19", "20", "21", "22", "23", "24", "25",
> "26", "27", "28", "29", "30", "31", "32", "33", "34", "35", "36",
> "37", "38", "39", "40", "41", "42", "43", "44", "45", "46", "47",
> "48", "49", "50", "51", "52", "53", "54", "55", "56", "57", "58",
> "59", "60", "61", "62", "63", "64", "65", "66", "67", "68", "69",
> "70", "71", "72"))
>
>
> # Modify the ?boxplot example
> bymedian <- with(DF, reorder(interaction(spray, op), count, median))
>
> > bymedian
>  [1] A.op1 A.op1 A.op1 A.op1 A.op1 A.op1 A.op2 A.op2 A.op2 A.op2 A.op2
> [12] A.op2 B.op1 B.op1 B.op1 B.op1 B.op1 B.op1 B.op2 B.op2 B.op2 B.op2
> [23] B.op2 B.op2 C.op1 C.op1 C.op1 C.op1 C.op1 C.op1 C.op2 C.op2 C.op2
> [34] C.op2 C.op2 C.op2 D.op1 D.op1 D.op1 D.op1 D.op1 D.op1 D.op2 D.op2
> [45] D.op2 D.op2 D.op2 D.op2 E.op1 E.op1 E.op1 E.op1 E.op1 E.op1 E.op2
> [56] E.op2 E.op2 E.op2 E.op2 E.op2 F.op1 F.op1 F.op1 F.op1 F.op1 F.op1
> [67] F.op2 F.op2 F.op2 F.op2 F.op2 F.op2
> attr(,"scores")
> A.op1 B.op1 C.op1 D.op1 E.op1 F.op1 A.op2 B.op2 C.op2 D.op2 E.op2 F.op2
>  13.0  15.0   1.5   4.5   4.0  15.0  15.5  17.0   1.5   5.0   2.5  18.5
> 12 Levels: C.op1 C.op2 E.op2 E.op1 D.op1 D.op2 A.op1 B.op1 ... F.op2
>
>
> boxplot(count ~ bymedian, data = DF,
>         xlab = "Interaction of spray and op", ylab = "Insect count",
>         main = "Modified InsectSprays Data", varwidth = TRUE,
>         col = "lightgray")
>
>
> Regards,
>
> Marc Schwartz
>
>

[[alternative HTML version deleted]]

```