[Ben Jann]

Follow up to #15: the format of the shape information in the excel file is called WKT ("Well-known text representation of geometry", see https://en.wikipedia.org/wiki/Well-k...on_of_geometry), so the program I provided should probably better be called wkt2dta or so. Currently the program only supports "MultiPolygon", but it could easily be extended to support points and lines as well. I'll think about generalizing the program and integrating it into the geoframe system.

[Matthew Williams]

Dear Prof. Ben Jann,

Thank you so much for the kind help in #15.

[Michael Duarte Goncalves]

Hi everyone,

I need your help in -geoplot-. My version of stata is 17.
Here are my questions, please:

1. First, I have some unmatched zipcodes from my "using" dataset, and I would like to match them to neighbouring zipcodes, if and only if the unmatched zipcodes contain more than 20 (>=20) households.
2. Second, in the map that I will show you afterwards, I would like to have the Canary Islands near to the Iberian Peninsula, but I don't know how.
3. Third, I would like to paint the white areas in my map into gray ones, since we do not have data on them.

Here is a dataex sorted of my unmatched zipcodes (from the using dataset):

* Example generated by -dataex-. For more info, type help dataex
clear
input str5 sp_zipcode_str long tag_zip_house_mun byte _merge
"08092"   1 2
"08190" 264 2
"08200"   2 2
"08220"   2 2
"08240"  15 2
"08400"  51 2
"10683"   1 2
"13197"   1 2
"15192"   6 2
"15407"   1 2
"15708"   1 2
"17092"   1 2
"17254"   1 2
"18849"   1 2
"18857"   1 2
"19019"   1 2
"21102"   1 2
"21458"   1 2
"21646"   1 2
"24400"  22 2
"26000"   1 2
end
label values _merge _merge
label def _merge 2 "Using only (2)", modify

tag_zip_house_mun corresponds to the number of observations in each zipcode.
Here is a dataex with the matched zipcodes:

* Example generated by -dataex-. For more info, type help dataex
clear
input str5 sp_zipcode_str byte _merge
"08100" 3
"08105" 3
"08106" 3
"08107" 3
"08110" 3
"08120" 3
"08130" 3
"08140" 3
"08148" 3
"08150" 3
"08160" 3
"08170" 3
"08172" 3
"08173" 3
"08174" 3
"08178" 3
"08180" 3
"08181" 3
"08182" 3
"08183" 3
"08184" 3
"08185" 3
"08186" 3
"08187" 3
"08188" 3
"08189" 3
"08191" 3
end
label values _merge _merge
label def _merge 3 "Matched (3)", modify

If needed, I can share the entire merge == 2 & merge == 3 data.

Finally, here is my code:

cd ${stata}/1_export_contratos/data

*** by zipcodes municipalities
use "export_contratos_original_date_cleaned.dta", clear

duplicates drop id, force

tostring sp_zipcode, gen(my_newzip) format(%07.0f)
gen sp_zipcode_str = substr(my_newzip,3, 2)
replace sp_zipcode_str = substr(my_newzip,3, 5)
sort sp_zipcode_twodigits
drop my_newzip

order id id_encod sp_zipcode sp_zipcode_str

egen tag_zip_house_mun = tag(id sp_zipcode_str) if !missing(powers_tariff2_less_15000w)
egen obs_by_zipcode_house_mun = total(tag_zip_house_mun), by(sp_zipcode_str)



sort sp_zipcode
collapse (count) tag_zip_house_mun, by(sp_zipcode_str)

tempfile house_info_munic
save `house_info_munic', replace

save "${stata}/1_export_contratos/data/mapping_data/house_info_collapse_municip.dta", replace


*******Map at the municipal level
clear all
cd ${export_geopath}

use "Spain_municip.dta", clear
gen sp_zipcode_str = COD_POSTAL
sort sp_zipcode_str
save Spain_municip.dta, replace



*1)Creation of Provincias limits
geoframe create Spain_provinc, replace
geoframe describe Spain_provinc    

*2)Creation of Municipal limits
geoframe create Spain_municip, replace
geoframe describe Spain_municip

*3)Some data cleaning, if needed
geoframe describe Spain_municip_shp

duplicates drop sp_zipcode_str, force

merge 1:1 sp_zipcode_str using `house_info_munic', keep(match master) nogenerate
save Spain_municip, replace


*4)Creation of CCAA limits
geoframe create Spain_auton
geoframe describe Spain_auton_shp

*5)Creation of collapse connexion
geoframe create "${stata}/1_export_contratos/data/mapping_data/house_info_collapse_municip", replace
geoframe describe house_info_collapse_municip


geoplot ///
    (area Spain_municip tag_zip_house_mun, color(reds)) ///
    (line Spain_auton, lwidth(0.05)) ///
    , tight legend(pos(9))

Please find below and enclosed the map that I mentioned above.
Could you please help me with that?

Your help is really appreciated!

Best,
Michael

[Ben Jann]

1. First, I have some unmatched zipcodes from my "using" dataset, and I would like to match them to neighbouring zipcodes, if and only if the unmatched zipcodes contain more than 20 (>=20) households.

Not really clear to me, what you want to do here.

2. Second, in the map that I will show you afterwards, I would like to have the Canary Islands near to the Iberian Peninsula, but I don't know how.

A common way to do this would be to modify the coordinates of the shape items belonging to the Canary Islands before plotting. However, you could also (mis)use the zoom() option in geoplot. This requires that you place the Canary Islands in a separate layer.

3. Third, I would like to paint the white areas in my map into gray ones, since we do not have data on them.

It is unclear to me why you have the white areas in the map, since there are also gray areas for municipalities without data. It seems that for the white areas there is no shape information in the municipalities data. Make sure that your data contains shape information on all municipalities.