fix ancient date format

docs/main.md

@@ -62,7 +62,9 @@ Curate metadata with a DB Browser (SQLite). Examples of modifying the BioSampleC
 - Annotate with meaningful metadata.
   - Add collection data, geographic location, host.
 
-### Geocode
+### Geospatial
+
+#### Geocoding
 
 Geographic location for samples is coded at the level of country and
 province/state in the format "Country:Province". Optional sub-province level
@@ -90,6 +92,12 @@ python workflow/scripts/geocode.py "Armenia:Shirak Province"
 > Shirak Province, Armenia
 > 40.918594 43.8403536
 
+#### Palladio
+
+```bash
+bash workflow/scripts/palladio.sh 2>&1 | tee results/metadata/sra/palladio_ancient.tsv
+```
+
 ## Genomic Alignment
 
 ### Modern Assembly Remote

workflow/scripts/geocode.py

@@ -4,11 +4,16 @@
 
 args = sys.argv
 place_name = args[1]
+DELIM = "\t"
 
 geolocator = Nominatim(user_agent="plague-phylogeography")
 country_address = place_name.split(":")[0]
 province_address = ":".join(place_name.split(":")[0:2])
 
+country_name = country_address
+province_name = place_name.split(":")[1:2][0]
+
+"""
 # Geocode at country level
 location = geolocator.geocode(country_address, language="en",)
 print(location.address)
@@ -19,3 +24,7 @@
     location = geolocator.geocode(province_address, language="en",)
     print(location.address)
     print(location.latitude, location.longitude)
+"""
+
+location = geolocator.geocode(province_address, language="en",)
+print(str(location.latitude) + "," + str(location.longitude))

workflow/scripts/palladio.sh

@@ -0,0 +1,39 @@
+#!/bin/bash
+
+# Output Format
+# 1. BioSampleAccession
+# 2. Strain
+# 3. Country
+# 4. Province
+# 6. BeginDate
+# 7. EndDate
+# 8. LatLon
+
+QUERY="SELECT
+  BioSampleAccession,
+	BioSampleStrain,
+  BioSampleGeographicLocation,
+  BioSampleCollectionDate
+FROM
+  BioSample
+WHERE
+  BioSampleComment LIKE '%KEEP%SRA%Ancient'";
+
+DELIM="\t"
+
+echo -e "Accession"$DELIM"Strain"$DELIM"Country"$DELIM"Province"$DELIM"Begin"$DELIM"End"$DELIM"LatLon"
+
+sqlite3 \
+  results/sqlite_db/yersinia_pestis_db.sqlite \
+  "$QUERY" | \
+  while read line;
+  do
+    acc=`echo $line | cut -d "|" -f 1`;
+		strain=`echo $line | cut -d "|" -f 2`;
+    country=`echo $line | cut -d "|" -f 3 | cut -d ":" -f 1`;
+    province=`echo $line | cut -d "|" -f 3 | cut -d ":" -f 2`;
+    begin=`echo $line | cut -d "|" -f 4 | cut -d ":" -f 1`;
+    end=`echo $line | cut -d "|" -f 4 | cut -d ":" -f 2`;
+		latlon=`python workflow/scripts/geocode.py "$country:$province"`;
+		echo -e	$acc$DELIM$strain$DELIM$country$DELIM$province$DELIM$begin$DELIM$end$DELIM$latlon;
+  done