Add colombian products

product/HS/Colombia_Prospedia/README.md

@@ -0,0 +1,3 @@
+This is basically HS92, with names generously translated by Ryan.
+
+[here](https://docs.google.com/spreadsheets/d/1yLhTZ6cV6aWLJkJmAQGRIEGKFSEeF_dwMtCA3YagOzI/edit#gid=0)

product/HS/Colombia_Prospedia/Tupfile

@@ -0,0 +1,3 @@
+CLEAN = PYTHONPATH=../../../ python2.7 -B clean.py
+
+: in/* |> $(CLEAN) |> out/products_colombia_prospedia.csv out/products_colombia_prospedia.dta

product/HS/Colombia_Prospedia/clean.py

@@ -0,0 +1,37 @@
+import pandas as pd
+
+from classification import (Hierarchy, repeated_table_to_parent_id_table,
+                            parent_code_table_to_parent_id_table,
+                            Classification)
+
+if __name__ == "__main__":
+    names = pd.read_table("./in/HS_hierarchy_master - Names.tsv", encoding="utf-8", dtype={"code": str})
+
+    hierarchy = pd.read_table("./in/HS_hierarchy_master - Hierarchy.tsv",
+                              encoding="utf-8",
+                              dtype={
+                                  "4digit": str,
+                                  "2digit": str,
+                                  "prosperia_section": str,
+                                  "section": str,
+                              })
+    hierarchy.columns = ["4digit_code", "2digit_code", "prosperia_section_code", "atlas_section"]
+    hierarchy["4digit_name"] = None
+    hierarchy["2digit_name"] = None
+    hierarchy["prosperia_section_name"] = None
+
+
+    h = Hierarchy(["prosperia_section", "2digit", "4digit"])
+    parent_code_table = repeated_table_to_parent_id_table(hierarchy, h)
+
+    parent_code_table = parent_code_table.merge(names, on=["code", "level"])
+
+    parent_id_table = parent_code_table_to_parent_id_table(parent_code_table, h)
+    parent_id_table.name = parent_id_table.name_en
+
+    from IPython import embed; embed()
+
+    c = Classification(parent_id_table, h)
+
+    c.to_csv("out/products_mexico_prospedia.csv")
+    c.to_stata("out/products_mexico_prospedia.dta")