chore(rust): move unique_counts to ops (#11963)

crates/polars-core/Cargo.toml

@@ -151,7 +151,6 @@ docs-selection = [
   "abs",
   "dataframe_arithmetic",
   "product",
-  "unique_counts",
   "describe",
   "chunked_ids",
   "partition_by",

crates/polars-core/src/series/ops/mod.rs

@@ -2,7 +2,6 @@ mod downcast;
 mod extend;
 mod null;
 mod to_list;
-mod unique;
 #[cfg(feature = "serde")]
 use serde::{Deserialize, Serialize};
 

crates/polars-ops/Cargo.toml

@@ -65,6 +65,7 @@ round_series = []
 is_first_distinct = []
 is_last_distinct = []
 is_unique = []
+unique_counts = []
 approx_unique = []
 fused = []
 cutqcut = ["dtype-categorical", "dtype-struct"]

crates/polars-ops/src/series/ops/mod.rs

@@ -44,6 +44,8 @@ mod round;
 mod search_sorted;
 #[cfg(feature = "to_dummies")]
 mod to_dummies;
+#[cfg(feature = "unique_counts")]
+mod unique;
 mod various;
 
 pub use approx_algo::*;
@@ -93,6 +95,8 @@ pub use round::*;
 pub use search_sorted::*;
 #[cfg(feature = "to_dummies")]
 pub use to_dummies::*;
+#[cfg(feature = "unique_counts")]
+pub use unique::*;
 pub use various::*;
 
 pub trait SeriesSealed {

crates/polars-ops/src/series/ops/unique.rs

@@ -0,0 +1,42 @@
+use std::hash::Hash;
+
+use polars_core::hashing::_HASHMAP_INIT_SIZE;
+use polars_core::prelude::*;
+use polars_core::utils::NoNull;
+
+fn unique_counts_helper<I, J>(items: I) -> IdxCa
+where
+    I: Iterator<Item = J>,
+    J: Hash + Eq,
+{
+    let mut map = PlIndexMap::with_capacity_and_hasher(_HASHMAP_INIT_SIZE, Default::default());
+    for item in items {
+        map.entry(item)
+            .and_modify(|cnt| {
+                *cnt += 1;
+            })
+            .or_insert(1 as IdxSize);
+    }
+    let out: NoNull<IdxCa> = map.into_values().collect();
+    out.into_inner()
+}
+
+/// Returns a count of the unique values in the order of appearance.
+pub fn unique_counts(s: &Series) -> PolarsResult<Series> {
+    if s.dtype().to_physical().is_numeric() {
+        if s.bit_repr_is_large() {
+            let ca = s.bit_repr_large();
+            Ok(unique_counts_helper(ca.into_iter()).into_series())
+        } else {
+            let ca = s.bit_repr_small();
+            Ok(unique_counts_helper(ca.into_iter()).into_series())
+        }
+    } else {
+        match s.dtype() {
+            DataType::Utf8 => Ok(unique_counts_helper(s.utf8().unwrap().into_iter()).into_series()),
+            dt => {
+                polars_bail!(opq = unique_counts, dt)
+            },
+        }
+    }
+}

crates/polars-plan/Cargo.toml

@@ -115,7 +115,7 @@ random = ["polars-core/random"]
 dynamic_group_by = ["polars-core/dynamic_group_by"]
 ewma = ["polars-ops/ewma"]
 dot_diagram = []
-unique_counts = ["polars-core/unique_counts"]
+unique_counts = ["polars-ops/unique_counts"]
 log = ["polars-ops/log"]
 chunked_ids = ["polars-core/chunked_ids"]
 list_to_struct = ["polars-ops/list_to_struct"]

crates/polars-plan/src/dsl/function_expr/dispatch.rs

@@ -54,7 +54,7 @@ pub(super) fn value_counts(s: &Series, sort: bool, parallel: bool) -> PolarsResu
 
 #[cfg(feature = "unique_counts")]
 pub(super) fn unique_counts(s: &Series) -> PolarsResult<Series> {
-    Ok(s.unique_counts().into_series())
+    polars_ops::prelude::unique_counts(s)
 }
 
 pub(super) fn backward_fill(s: &Series, limit: FillNullLimit) -> PolarsResult<Series> {

crates/polars/Cargo.toml

@@ -156,7 +156,7 @@ ewma = ["polars-ops/ewma", "polars-lazy?/ewma"]
 dot_diagram = ["polars-lazy?/dot_diagram"]
 dataframe_arithmetic = ["polars-core/dataframe_arithmetic"]
 product = ["polars-core/product"]
-unique_counts = ["polars-core/unique_counts", "polars-lazy?/unique_counts"]
+unique_counts = ["polars-ops/unique_counts", "polars-lazy?/unique_counts"]
 log = ["polars-ops/log", "polars-lazy?/log"]
 partition_by = ["polars-core/partition_by"]
 semi_anti_join = ["polars-lazy?/semi_anti_join", "polars-ops/semi_anti_join", "polars-sql?/semi_anti_join"]