[email protected]
> I am having trouble with a programme I wrote for specific
> data management
> purpose. I have 271 farmers who may combine at most 3 types
> of crop c1 c2 c3.
> The data is in long format and therefore ci records in
> column the type of
> crop per plot owned by the farmer and in row ci show the
> mixture of crop by
> the farmer in each of the plot. A lot of missing values
> appear of course in
> the data.
> My first objective was therefore to create dummies ci_di
> per crop (maximum of
> 46 types of crop) and this is done through the first
> section of my do file
> below but in my attempt to pass a label value that identify
> each type of crop
> I got the error : label beans already defined
> r(110);
> I actually used some resources of Statalist (posted by Nick
> but on different
> issues) to automate the labelling.
> *--------STEP1: tab, generating dummies per crop and
> labelling their values-
> version 7.0
> use ctypes, clear
> program define compo
> foreach var of local 0{
> sort id
> tab `var', gen(`var'_d)
>
> *below the local cropnames is a sample of the 46 types of
> crops I have. and
> *at the end of the program I pasted the error generated
>
> local cropnames "beans cassava maize rice watermelon
> macaxeira maxixe melon
> raizman tomato jambu goma cucumber Cheiroverde sweetpotato
> tabaco Jerimum
> PimentaCheiro Pimenta Doce Beringela"
>
> local codes "beans cassava maize rice watermelon macaxeira
> maxixe melon
> raizman tomato jambu goma cucumber Cheiroverde sweetpotato
> tabaco Jerimum
> PimentaCheiro Pimenta Doce Beringela"
>
> local i = 1
> foreach x of local cropnames {
> local y : word `i' of `codes'
> lab define `x' 1 "`y'" 0 "no"
> capture forvalues j=1/46{
> label val `var'_d`j' `x'
> sum area if `var'_d`j'==1
> }
> local i = `i' + 1
> }
> end
> local area "c1 c2 c3"
> foreach k of local area {
> compo `k'
> }
> program drop compo
> exit
> label beans already defined
> r(110);
I think this is solved by
lab def `x' 1 "`y'" 0 "no", modify
> The next section of the do file simply collapse by crop
> type to get the mean
> value of area croped. Here also I got more than hundred
> data files. Do I have
> to go through each file to display the output of interest
> (mean value of area
> croped) or is there any way I can gather these results all together?
> During the survey the farmer could not know the portion of
> land allocated to
> each crop and one methodological question is to know if it
> is a good idea to
> simply divide the total land by the number of crops
> cultivated on the plot.
> I would appreciate any help from the list.
> *------------------STEP 2: mean value of the area croped by
> crop type.
> *Everything works fine but how to put together the outputs
> generated by
> *collapse--------------------------
>
> version 7.0
> program define compo2
> foreach var of local 0 {
> use ctypes, clear
> sort id
> foreach v of var * {
> local l`v' : variable label `v'
> if `"`l`v''"' == "" {
> local l`v' "`v'"
> }
> }
> preserve
> capture forvalues i = 1/46 {
> use ctypes, clear
> collapse area if `var'_d`i'==1
> foreach v of var * {
> label var `v' "mean value of `l`v'' `var'_d`i'"
> }
> keep id plot area fareacrop
> save f`var'`i', replace
> }
> restore
> }
> end
> local area "c1 c2 c3"
> foreach j of local area {
> compo2 `j'
> }
> program drop compo2
> exit
I suspect the easiest way to get the mean
value is through -summarize-: you don't
need to -collapse-.
Nick
[email protected]
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
