U ꥡco@s"ddlZddlZddlZddlZddlZddlZddlmZddl m Z ddl m Z ddl m Z ddlmZddlmZe ZGd d d eeZGd d d eZejd)ddZejd*ddZejd+ddZejd,ddZejddZddZdd Zd!d"Zd#d$Zd%d&Z d'd(ZdS)-N)curried)Callable)sanitize_dataframe)sanitize_geo_interface)AltairDeprecationWarning)PluginRegistryc@s.eZdZddiZeddZejddZdS)DataTransformerRegistryconsolidate_datasetsTcCs |jdSNr _global_settings)selfr5/tmp/pip-unpacked-wheel-y9_o96ar/altair/utils/data.pyr sz,DataTransformerRegistry.consolidate_datasetscCs||jd<dSr r )rvaluerrrr sN)__name__ __module__ __qualname__r propertyr setterrrrrr s  r c@seZdZdZdS) MaxRowsErrorz+Raised when a data model has too many rows.N)rrr__doc__rrrrr6srcCst|t|dr4|jddkr,|jd}qf|j}n2t|tjrF|}n t|trfd|krb|d}n|S|dk rt||krtd ||S)zzRaise MaxRowsError if the data model has more than max_rows. If max_rows is None, then do not perform any check. __geo_interface__typeZFeatureCollectionfeaturesvaluesNzThe number of rows in your dataset is greater than the maximum allowed ({}). For information on how to plot larger datasets in Altair, see the documentation) check_data_typehasattrr isinstancepd DataFramedictlenrformat)dataZmax_rowsrrrr limit_rows<s$     r'cCslt|t|tjr"|j||dSt|trhd|krh|d}|rD|nt|t|}t||}d|iSdS)zBReduce the size of the data model by sampling without replacement.)nfracrN) rr r!r"sampler#intr$random)r&r(r)rrrrr*Ys   r* altair-datajson{prefix}-{hash}.{extension}c CsXt|}t|}|j|||d}t|d}||W5QRXtj||ddidS)zQ Write the data model to a .json file and return a url based data model. prefixhash extensionwrr.urlr%)_data_to_json_string_compute_data_hashr%openwriteospathjoin)r&r2r4filenameurlpathZ data_json data_hashfrrrto_jsongs  rCcsvc CsXt|}t|}|j|||d}t|d}||W5QRXtj||ddidS)zFWrite the data model to a .csv file and return a url based data model.r1r5rrDr6)_data_to_csv_stringr9r%r:r;r<r=r>)r&r2r4r?r@Zdata_csvrArBrrrto_csvzs  rFcCs~t|t|dr8t|tjr&t|}t|j}d|iSt|tjr\t|}d|jddiSt|t rzd|krvt d|SdS)z0Replace a DataFrame by a data model with values.rrrecords)orient.values expected in data dict, but not present.N) rrr r!r"rrrZto_dictr#KeyErrorr&rrr to_valuess     rLcCs0t|ttjfs,t|ds,tdt|dS)z-Raise if the data is not a dict or DataFrame.rzBExpected dict, DataFrame or a __geo_interface__ attribute, got: {}N)r r#r!r"r TypeErrorr%rrKrrrrsrcCst|S)N)hashlibmd5encode hexdigest)Zdata_strrrrr9sr9cCst|t|dr:t|tjr&t|}t|j}t |St|tjr\t|}|j dddSt|t rd|krvt dtj |dddSt d d S) z5Return a JSON string representation of the input datarrG)rHZdouble_precisionrrIT) sort_keyszBto_json only works with data expressed as a DataFrame or as a dictN)rrr r!r"rrrr.dumpsrCr#rJNotImplementedErrorrKrrrr8s       r8cCszt|t|drtdnZt|tjrsL