R help - Aug 2008 - Updating a list.

I have a list that is generated from the resape package function 'cast'.
It consists of three columns, Sku, DayOfYear, variable it is generated like:

r2007 <- cast(m2008, DayOfYear ~ variable | Sku, sum)

Now DayOfYear can range from 1:365 but there are not necessarily that many rows
in the list. What I want to do is make every row in the list of lenght 365 and
have the values correspond to the sum of the DayOfYear in the list if present
and 0 if not. For example the first item in the list looks like:

$`100026`
   DayOfYear Quantity
1          3        1
2         30        1
3        149        1
4        156        1
5          3         1

Notice that there are not entires for each 'DayOfYear'. I would like to
replace list list maintaining the same data structure with something like:

$`100026`
   DayOfYear Quantity
1          1        0
2          2        0
3          3        2
4          4        0

Where the Quantity column is the sum of the Quantity column in the previous list
with the same 'Sku'. Notice there are no entries for 1,2 and 4 so those
columns are zero and in the small example there are only two columns for
DayOfYear 3, so the sum is 2. If it helps I have included hints to the data
structure that I am working with below.

The parts that are particularly hard for me is how I would simultaneously remove
and replace a list and conditionally sum only if a value exists.

Thank you.

Kevin

dput(r007[1])

structure(list("100026" = structure(list(DayOfYear = c(3L, 30L, 
149L, 156L, 165L, 171L, 187L, 198L, 212L, 215L, 216L, 218L, 221L, 
224L, 226L, 227L), Quantity = c(1L, 1L, 1L, 1L, 2L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L)), .Names = c("DayOfYear",
"Quantity"
), row.names = c(NA, 16L), class = c("cast_df",
"data.frame"), idvars = "DayOfYear", rdimnames = list(
    structure(list(DayOfYear = c(3L, 30L, 149L, 156L, 165L, 171L, 
    187L, 198L, 212L, 215L, 216L, 218L, 221L, 224L, 226L, 227L
    )), .Names = "DayOfYear", row.names = c("3",
"30", "149",
    "156", "165", "171", "187",
"198", "212", "215", "216",
"218",
    "221", "224", "226", "227"), class =
"data.frame"), structure(list(
        variable = structure(1L, .Label = "Quantity", class =
"factor")), .Names = "variable", row.names =
"Quantity", class = "data.frame")))), .Names =
"100026")
> str(r2007[1])
List of 1
 $ 100009:List of 2
 $ DayOfYear: int [1:4] 66 128 137 193
 $ Quantity : int [1:4] 1 1 1 1
 - attr(*, "row.names")= int [1:4] 1 2 3 4
 - attr(*, "idvars")= chr "DayOfYear"
 - attr(*, "rdimnames")=List of 2
  ..$ :'data.frame':    4 obs. of  1 variable:
  .. ..$ DayOfYear: int [1:4] 66 128 137 193
  ..$ :'data.frame':    1 obs. of  1 variable:
  .. ..$ variable: Factor w/ 1 level "Quantity": 1

On Wed, Aug 27, 2008 at 6:18 PM,  <rkevinburton at charter.net>
wrote:
> I have a list that is generated from the resape package function
'cast'. It consists of three columns, Sku, DayOfYear, variable it is
generated like:
>
> r2007 <- cast(m2008, DayOfYear ~ variable | Sku, sum)
>
> Now DayOfYear can range from 1:365 but there are not necessarily that many
rows in the list. What I want to do is make every row in the list of lenght 365
and have the values correspond to the sum of the DayOfYear in the list if
present and 0 if not. For example the first item in the list looks like:
>
> $`100026`
>   DayOfYear Quantity
> 1          3        1
> 2         30        1
> 3        149        1
> 4        156        1
> 5          3         1
>
> Notice that there are not entires for each 'DayOfYear'. I would
like to replace list list maintaining the same data structure with something
like:
>
> $`100026`
>   DayOfYear Quantity
> 1          1        0
> 2          2        0
> 3          3        2
> 4          4        0
>
Try this:

m2008$DayOfYear <- factor(m2008$DayOfYear, levels = 1:365)
r2007 <- cast(m2008, DayOfYear ~ variable | Sku, sum, fill = 0)

Hadley

-- 
http://had.co.nz/