o * i"@sdZddlZddlZddlZddlZddlZddlm Z gZ gdZ dZ dZ GdddZGd d d Zdadadadad d Zd1ddZe dddddddZejeddZejeddZe dddddddZddZe dddddddZe dddddd d!Zd"d#Ze dddddd$d%Ze dddddd&d'Z e dddddd(d)Z!e dddddd*d+Z"d,d-Z#e dddddd.d/Z$e%d0kre#dSdS)2aD Movielens 1-M dataset. Movielens 1-M dataset contains 1 million ratings from 6000 users on 4000 movies, which was collected by GroupLens Research. This module will download Movielens 1-M dataset from http://files.grouplens.org/datasets/movielens/ml-1m.zip and parse training set and test set into paddle reader creators. N) deprecated)#-28z3https://dataset.bj.bcebos.com/movielens%2Fml-1m.zipZ c4d9eecfca2ab87c1945afe126590906c@0eZdZdZddZddZddZdd Zd S) MovieInfozM Movie id, title and categories information are stored in MovieInfo. cCst||_||_||_dSN)intindex categoriestitle)selfrrrrd/home/app/PaddleOCR-VL-test/.venv_paddleocr/lib/python3.10/site-packages/paddle/dataset/movielens.py__init__0s  zMovieInfo.__init__cCs(|jdd|jDdd|jDgS)z/ Get information from a movie. cSsg|]}t|qSr)CATEGORIES_DICT).0crrr ;sz#MovieInfo.value..cSsg|]}t|qSr)MOVIE_TITLE_DICTlower)rwrrrr<s)rrrsplitrrrrvalue5szMovieInfo.valuecCsd|jd|jd|jdS)Nz)rrrrrrr__str__?szMovieInfo.__str__cCs|Sr )r rrrr__repr__BzMovieInfo.__repr__N__name__ __module__ __qualname____doc__rrr r!rrrrr +s  r c@r ) UserInfozK User id, gender, age, and job information are stored in UserInfo. cCs2t||_|dk|_tt||_t||_dS)NM)r ris_male age_tableagejob_id)rrgenderr,r-rrrrKs  zUserInfo.__init__cCs|j|jrdnd|j|jgS)z. Get information from a user. rr)rr*r,r-rrrrrQszUserInfo.valuec Cs6|jrdnd}d|jd|dt|jd|jd S)Nr)Fz Please use new dataset API which supports paddle.io.DataLoader)ZsinceZ update_tolevelreasonc s fddS)NcstdiS)Nr)rarkwargsrrsz$__reader_creator__..rrdrrdr__reader_creator__s rg)r_TcC ttS)z% Get movie title dictionary. )rXrrrrrget_movie_title_dict ricC|j|jkr|S|Sr )rabrrr__max_index_info__ rocCtttttjS)z, Get the maximum value of movie id. )rX functoolsreducerolistr?valuesrrrrr max_movie_id rvcCrq)z+ Get the maximum value of user id. )rXrrrsrortrOrurrrrr max_user_idrwrxcCrkr )r-rlrrr__max_job_id_impl__rprycCrq)z* Get the maximum value of job id. )rXrrrsryrtrOrur-rrrr max_job_ids   rzcCrh)z* Get movie categories dictionary. )rXrrrrrmovie_categoriesrjr{cCrh)z# Get user info dictionary. )rXrOrrrr user_info rjr|cCrh)z$ Get movie info dictionary. )rXr?rrrr movie_inforjr}cCs:ttD]\}}qttD]\}}qt||dSr )rItraintestprint)Z train_countrWZ test_countrrrunittest(s rcCstjjtdtdS)Nr1)r9r:r;r<r=r>rrrrfetch1sr__main__)rrYF)&r'rrr@rBnumpyr[Zpaddle.dataset.commonr9Z paddle.utilsr__all__r+r=r>r r(r?rrrOrXrargpartialr~rrirorvrxryrzr{r|r}rrr$rrrrs    /