ltc-code 0.1.7__tar.gz → 0.1.8__tar.gz

{ltc_code-0.1.7 → ltc_code-0.1.8}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: ltc-code
-Version: 0.1.7
+Version: 0.1.8
 Summary: Add your description here
 Requires-Python: >=3.9
 Description-Content-Type: text/markdown

{ltc_code-0.1.7 → ltc_code-0.1.8}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "ltc-code"
-version = "0.1.7"
+version = "0.1.8"
 description = "Add your description here"
 readme = "README.md"
 requires-python = ">=3.9"

{ltc_code-0.1.7 → ltc_code-0.1.8}/src/ltc_code/may27.py

@@ -1424,9 +1424,6 @@ def lookup_sid_cepr(
 ####################################################################################
 
 
-
-
-
 def _parse_dob_expr(col: str) -> pl.Expr:
     return pl.coalesce(
         [
@@ -1437,6 +1434,14 @@ def _parse_dob_expr(col: str) -> pl.Expr:
     )
 
 
+def _first_word_expr(col: str) -> pl.Expr:
+    return pl.col(col).cast(pl.String).str.split(" ").list.first()
+
+
+def _second_word_expr(col: str) -> pl.Expr:
+    return pl.col(col).cast(pl.String).str.split(" ").list.get(1, null_on_oob=True)
+
+
 def _build_lookup(
     census: pl.DataFrame,
     *,
@@ -1456,13 +1461,11 @@ def _build_lookup(
         .drop_nulls(["_fname_key", "_lname_key", "_dob_key", "sid_cepr"])
         .group_by(["_fname_key", "_lname_key", "_dob_key"])
         .agg(pl.col("sid_cepr").unique().alias("_sids"))
-        .with_columns(pl.col("_sids").list.len().alias("_sid_count"))
-        .filter(pl.col("_sid_count") == 1)
         .select(
             "_fname_key",
             "_lname_key",
             "_dob_key",
-            pl.col("_sids").list.first().alias("sid_cepr"),
+            pl.col("_sids").list.sort().list.first().alias("sid_cepr"),
         )
     )
 
@@ -1603,6 +1606,30 @@ def build_census_lookups(*, cmo_name: str) -> dict[str, pl.DataFrame]:
         label="right lname + suffix no space",
     )
 
+    lookups["fname_first_word"] = _build_lookup(
+        census,
+        fname_expr=_first_word_expr("fname"),
+        lname_expr=pl.col("lname"),
+        dob_col="dob",
+        label="right fname first word",
+    )
+
+    lookups["lname_first_word"] = _build_lookup(
+        census,
+        fname_expr=pl.col("fname"),
+        lname_expr=_first_word_expr("lname"),
+        dob_col="dob",
+        label="right lname first word",
+    )
+
+    lookups["lname_second_word"] = _build_lookup(
+        census,
+        fname_expr=pl.col("fname"),
+        lname_expr=_second_word_expr("lname"),
+        dob_col="dob",
+        label="right lname second word",
+    )
+
     lookups["dob_imp"] = _build_lookup(
         census,
         fname_expr=pl.col("fname"),
@@ -1659,7 +1686,6 @@ def _run_match_stage(
     unmatched = stage.filter(pl.col("sid_cepr").is_null()).drop("sid_cepr")
 
     print(f"{label}: matched {len(matched):,}/{before:,}")
-
     return matched, unmatched
 
 
@@ -1710,6 +1736,9 @@ def lookup_sid_cepr(
         ("left exact -> right fname + suffix", lookups["suffix_fname"]),
         ("left exact -> right fname + suffix no space", lookups["suffix_fname_nospace"]),
         ("left exact -> right lname + suffix no space", lookups["suffix_lname_nospace"]),
+        ("left exact -> right fname first word", lookups["fname_first_word"]),
+        ("left exact -> right lname first word", lookups["lname_first_word"]),
+        ("left exact -> right lname second word", lookups["lname_second_word"]),
     ]:
         matched, unmatched = _run_match_stage(
             unmatched,
@@ -1721,6 +1750,33 @@ def lookup_sid_cepr(
         )
         matched_frames.append(matched)
 
+    for label, fname_expr, lname_expr in [
+        (
+            "left fname first word -> right exact",
+            _first_word_expr(f"{cols['fname']}_clean"),
+            pl.col(f"{cols['lname']}_clean"),
+        ),
+        (
+            "left lname first word -> right exact",
+            pl.col(f"{cols['fname']}_clean"),
+            _first_word_expr(f"{cols['lname']}_clean"),
+        ),
+        (
+            "left lname second word -> right exact",
+            pl.col(f"{cols['fname']}_clean"),
+            _second_word_expr(f"{cols['lname']}_clean"),
+        ),
+    ]:
+        matched, unmatched = _run_match_stage(
+            unmatched,
+            lookup=lookups["exact"],
+            fname_expr=fname_expr,
+            lname_expr=lname_expr,
+            dob_expr=pl.col(f"{cols['dob']}_clean"),
+            label=label,
+        )
+        matched_frames.append(matched)
+
     if "mname" in cols:
         unmatched = unmatched.with_columns(*clean_other_name(cols["mname"]))
 
@@ -1838,7 +1894,6 @@ def lookup_sid_cepr(
     print(f"final matched: {result['sid_cepr'].is_not_null().sum():,}/{len(result):,}")
 
     return result
-#
 
 # EXAMPLE USAGE