The observer deployment data from the 2014 Afghanistan runoff elections was provided in various forms (excel, word) and saved as xlsx files (so the work is visible) and then csvs with minimal editing. That data is all read into R and joined with the updated polling center data to create the TileMill source data (as a csv).
For manual edits, I added the new name information into rosetta_fix.csv and the new name/spatial information (calculated from iec_runoff_polling_centers_en_jun2014-geojson_cl.csv") to district_centroid.csv.
All the extra lat/long calculations were saved as extracentroidcalc.xlsx.
FEFA didn't have IEC ID so I sorted the information by province then district and put it next to a similar sorted list from the rosetta stone file (province, district, plus IEC ID). TManually went through the whole list to do name matching
For Helmand-Marja, added IEC ID 3014 to match the polling station information (later adding lat/long based on the average of the same for all the Marja polling stations, added as AGCHO ID 9998)
Same for Delaram district in Nimroz, ID 3406 (it's apparently transfered provinces a few times, AGCHO 9999)
For Dari Suf Bala and Dari Suf Payan, I adjusted the rosetta fix csv so that IEC ID 2006 corresponds to AGCHO ID 1904 and 1905. This was reflected on the first observer deployment map but not totally sure how the matching was managed then. Because of this, the FEFA data has to be joined on AGCHO ID (since the IEC ID isn't unique)
All the numbered districts in Kabul province were assigned to Kabul district (and aggregated)
Added these places to the rosetta and district centroid files:
- Meerzaka, Paktya IEC 1313 AGCHO 9997
- Chenarto, Uruzgan IEC 2506 AGCHO 9996
- Dand, Kandahar IEC 2717 AGCHO 9995
- Takhta Pul, Kandahar IEC 2718 AGCHO 9994
Added these places to the rosetta and district centroid files:
- Abshar, Panjshir IEC 0808 AGCHO 9993
Performed a manual merge similar to the FEFA data to match province/district name with IEC/AGCHO IDs
Combined the observer numbers for Baghlan and Baghlani-Jadid (since the former was absorbed into the latter in 2005)
Combined the observer numbers for Sia Gird and Ghorband under Syahgirdi Gho (Syahgirdi Ghorband; IEC/AGCHO 0307)
Added these places to the rosetta and district centroid files:
- Laja Mangal, Paktya IEC 1312 AGCHO 9992
For mapping the results to the 2009 district shapefile (allowing me to bring in the population data) I did the following: