reshape2 - Convert from long wide data fromat to long data format in R with multiple depth headers -

maybe can't done because have data in bad format, i'm trying "kind of" wide format long format. if modify data in excel first, know how melt, have lot of data, , i'm hoping not have modify hand first.

basically have headers have depth have multiple descriptors each column of sales. account, brand, variety , tier (and maybe others depending on how goes)

            account             brand             variety             tier dates    1/1/2011    sales xxx 2/1/2011    sales xxx 

and format need in be

date account brand variety tier sales 

i don't understand how convert using reshape or melt though. suggestions, or have hand?

edit

actual data looks (names changed protect innocent)

account   account account   account   account   account b brand     brand   brand         brand b         brand b         brand c variety   cab           chard           merlot          pinot grigio    pinot noir  tier     tier 2    tier 3          tier 2          tier 1          tier 3 date                     6/1/20     35            47              35               28               -    7/1/2011   12            17              12               12              84                8/1/2011   -             -               -                -                -    9/1/2011   7             7           7            7               84  10/1/2011 28             28              28              28              112  11/1/2011                                            168  12/1/2011  7             7           7            7              224  1/1/2012                 14                              236  

if massage in excel this

    date    compilation_red compilation_white   oak_cab oak_chard 1   11/1/2012   351 140 183 190 2   12/1/2012   191 99  120 92 3   1/1/2013    96  56  87  51 4   2/1/2013    30  30  84  61 5   3/1/2013    96  70  100 57 6   4/1/2013    60  38  52  22 7   5/1/2013    12  17  65  49 8   6/1/2013    9   7   28  21 9   7/1/2013    16  13  23  22 10  8/1/2013    26  29  40  33 11  9/1/2013    14  4   30  23 12  10/1/2013   17  4   47  31 13  11/1/2013   23  5   63  35 14  12/1/2013   25  3   70  41 15  1/1/2014    44  1   88  41 16  2/1/2014    28  7   57  43 

then know how use melt long format

myitem=melt(account, id.vars="date", variable.name="variety", value.name="sales") 

i don't understand how deal headers, , them long format.

also, don't understand why table looks fine when put in, mashed in preview.