{"links":{},"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00222652","sets":["1164:2240:10902:11046"]},"path":["11046"],"owner":"44499","recid":"222652","title":["Ducks: DataFrame with Compiler"],"pubdate":{"attribute_name":"公開日","attribute_value":"2022-11-24"},"_buckets":{"deposit":"0f8232a0-828b-440d-aa8e-7dea8aeec512"},"_deposit":{"id":"222652","pid":{"type":"depid","value":"222652","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"Ducks: DataFrame with Compiler","author_link":["584416","584415","584413","584411","584414","584412"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"Ducks: DataFrame with Compiler"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"プログラミング","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2022-11-24","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"NECデジタルテクノロジー開発研究所"},{"subitem_text_value":"NECデジタルテクノロジー開発研究所"},{"subitem_text_value":"NECデジタルテクノロジー開発研究所"},{"subitem_text_value":"NECデジタルテクノロジー開発研究所"},{"subitem_text_value":"NECデジタルテクノロジー開発研究所"},{"subitem_text_value":"NECデジタルテクノロジー開発研究所"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"item_publisher":{"attribute_name":"出版者","attribute_value_mlt":[{"subitem_publisher":"情報処理学会","subitem_publisher_language":"ja"}]},"publish_status":"0","weko_shared_id":-1,"item_file_price":{"attribute_name":"Billing file","attribute_type":"file","attribute_value_mlt":[{"url":{"url":"https://ipsj.ixsq.nii.ac.jp/record/222652/files/IPSJ-HPC22187009.pdf","label":"IPSJ-HPC22187009.pdf"},"date":[{"dateType":"Available","dateValue":"2024-11-24"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-HPC22187009.pdf","filesize":[{"value":"1.0 MB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"660","billingrole":"5"},{"tax":["include_tax"],"price":"330","billingrole":"6"},{"tax":["include_tax"],"price":"0","billingrole":"14"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"84489d3f-a0ce-4fa9-8242-7412351650f8","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2022 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"石坂, 一久"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"大野, 善之"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Sourav, Saha"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"大道, 修"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"小寺, 雅司"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"荒木, 拓也"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN10463942","subitem_source_identifier_type":"NCID"}]},"item_4_textarea_12":{"attribute_name":"Notice","attribute_value_mlt":[{"subitem_textarea_value":"SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc."}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_18gh","resourcetype":"technical report"}]},"item_4_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"2188-8841","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"データフレームコンパイラを搭載し利便性と高速性の両方を実現するデータフレーム用 Python ライブラリ Ducks を紹介する．データフレームは，github で 35K star を誇る Pandas に代表されるように，データ分析や前処理に広く用いられているが，データ量の増大や分析の複雑化により高速化が求められている．Ducks は HPC で培われてきた実行時コンパイル技術を用いることで，ライブラリ呼び出しを直接実行するのではなく，データフーム用の中間言語（IR）を生成して遅延実行を行う方式を採用している．これにより API とその実行を分離し，Pandas 互換の API を提供しながら，IR 上でのドメイン特化の最適化，ターゲットプラットフォームに最適化されたバックエンドによる IR 実行により高速性も実現する．ベンチマーク集である TPCx-BB，TPC-H に含まれる 45 種類のデータ前処理・分析処理を用いた CPU 上での評価では，ライブラリを選択する impor t文の変更のみで，Pandas に対して最大 17 倍，平均 5.8 倍の性能向上を得ることができ，本方式の有効性を確認した．","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"6","bibliographic_titles":[{"bibliographic_title":"研究報告ハイパフォーマンスコンピューティング（HPC）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2022-11-24","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"9","bibliographicVolumeNumber":"2022-HPC-187"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"created":"2025-01-19T01:22:35.508348+00:00","updated":"2025-01-19T13:39:53.946649+00:00","id":222652}