Skip to content

PERF: restore kh_get before kh_put in hashtable operations #65026

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
mroeschke merged 1 commit into from
Apr 6, 2026
+31 −24
Merged

PERF: restore kh_get before kh_put in hashtable operations #65026

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
26 changes: 15 additions & 11 deletions pandas/_libs/hashtable_class_helper.pxi.in
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -757,10 +757,11 @@ cdef class {{name}}HashTable(HashTable):
append_data_uint8(rmd, 1)
continue

k = kh_put_{{dtype}}(self.table, val, &ret)
k = kh_get_{{dtype}}(self.table, val)

if ret != 0:
if k == self.table.n_buckets:
# k hasn't been seen yet
k = kh_put_{{dtype}}(self.table, val, &ret)

if needs_resize(ud.size, ud.capacity):
with gil:
Expand Down Expand Up @@ -898,8 +899,12 @@ cdef class {{name}}HashTable(HashTable):
labels[i] = -1
continue

k = kh_put_{{dtype}}(self.table, val, &ret)
if ret != 0:
k = kh_get_{{dtype}}(self.table, val)
if k != self.table.n_buckets:
idx = self.table.vals[k]
labels[i] = idx
else:
k = kh_put_{{dtype}}(self.table, val, &ret)
self.table.vals[k] = count

if needs_resize(ud.size, ud.capacity):
Expand All @@ -908,9 +913,6 @@ cdef class {{name}}HashTable(HashTable):
append_data_{{dtype}}(ud, val)
labels[i] = count
count += 1
else:
idx = self.table.vals[k]
labels[i] = idx

arr_uniques = uniques.to_array()

Expand Down Expand Up @@ -1219,9 +1221,10 @@ cdef class StringHashTable(HashTable):
continue

v = vecs[i]
k = kh_put_str(self.table, v, &ret)
if ret != 0:
k = kh_get_str(self.table, v)
if k == self.table.n_buckets:
# k hasn't been seen yet
k = kh_put_str(self.table, v, &ret)
uindexer[count] = i
if return_inverse:
self.table.vals[k] = count
Expand Down Expand Up @@ -1491,9 +1494,10 @@ cdef class PyObjectHashTable(HashTable):
labels[i] = na_sentinel
continue

k = kh_put_pymap(self.table, <PyObject*>val, &ret)
if ret != 0:
k = kh_get_pymap(self.table, <PyObject*>val)
if k == self.table.n_buckets:
# k hasn't been seen yet
k = kh_put_pymap(self.table, <PyObject*>val, &ret)
uniques.append(val)
if return_inverse:
self.table.vals[k] = count
Expand Down
29 changes: 16 additions & 13 deletions pandas/_libs/hashtable_func_helper.pxi.in
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -66,12 +66,13 @@ cdef value_count_{{dtype}}(const {{dtype}}_t[:] values, bint dropna, const uint8
for i in range(n):
val = values[i]
if not dropna or not checknull(val):
k = kh_put_{{ttype}}(table, {{to_c_type}}val, &ret)
if ret != 0:
k = kh_get_{{ttype}}(table, {{to_c_type}}val)
if k != table.n_buckets:
table.vals[k] += 1
else:
k = kh_put_{{ttype}}(table, {{to_c_type}}val, &ret)
table.vals[k] = 1
result_keys.append(val)
else:
table.vals[k] += 1
{{else}}
kh_resize_{{ttype}}(table, n)

Expand All @@ -89,12 +90,13 @@ cdef value_count_{{dtype}}(const {{dtype}}_t[:] values, bint dropna, const uint8
if uses_mask and isna_entry:
na_counter += 1
else:
k = kh_put_{{ttype}}(table, val, &ret)
if ret != 0:
k = kh_get_{{ttype}}(table, val)
if k != table.n_buckets:
table.vals[k] += 1
else:
k = kh_put_{{ttype}}(table, val, &ret)
table.vals[k] = 1
result_keys.append(val)
else:
table.vals[k] += 1
{{endif}}

# collect counts in the order corresponding to result_keys:
Expand Down Expand Up @@ -191,13 +193,14 @@ cdef duplicated_{{dtype}}(const {{dtype}}_t[:] values, object keep='first', cons

else:
value = {{to_c_type}}(values[i])
k = kh_put_{{ttype}}(table, value, &ret)
if ret != 0:
table.vals[k] = i
out[i] = 0
else:
k = kh_get_{{ttype}}(table, value)
if k != table.n_buckets:
out[table.vals[k]] = 1
out[i] = 1
else:
k = kh_put_{{ttype}}(table, value, &ret)
table.vals[k] = i
out[i] = 0

kh_destroy_{{ttype}}(table)
return out
Expand Down
Loading