0001 #
0002 # Licensed to the Apache Software Foundation (ASF) under one or more
0003 # contributor license agreements. See the NOTICE file distributed with
0004 # this work for additional information regarding copyright ownership.
0005 # The ASF licenses this file to You under the Apache License, Version 2.0
0006 # (the "License"); you may not use this file except in compliance with
0007 # the License. You may obtain a copy of the License at
0008 #
0009 # http://www.apache.org/licenses/LICENSE-2.0
0010 #
0011 # Unless required by applicable law or agreed to in writing, software
0012 # distributed under the License is distributed on an "AS IS" BASIS,
0013 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
0014 # See the License for the specific language governing permissions and
0015 # limitations under the License.
0016 #
0017
0018 # Imports from base R
0019 # Do not include stats:: "rpois", "runif" - causes error at runtime
0020 importFrom("methods", "setGeneric", "setMethod", "setOldClass")
0021 importFrom("methods", "is", "new", "signature", "show")
0022 importFrom("stats", "gaussian", "setNames")
0023 importFrom("utils", "download.file", "object.size", "packageVersion", "tail", "untar")
0024
0025 # Disable native libraries till we figure out how to package it
0026 # See SPARKR-7839
0027 #useDynLib(SparkR, stringHashCode)
0028
0029 # S3 methods exported
0030 export("sparkR.session")
0031 export("sparkR.init")
0032 export("sparkR.session.stop")
0033 export("sparkR.stop")
0034 export("sparkR.conf")
0035 export("sparkR.version")
0036 export("sparkR.uiWebUrl")
0037 export("print.jobj")
0038
0039 export("sparkR.newJObject")
0040 export("sparkR.callJMethod")
0041 export("sparkR.callJStatic")
0042
0043 export("install.spark")
0044
0045 export("sparkRSQL.init",
0046 "sparkRHive.init")
0047
0048 # MLlib integration
0049 exportMethods("glm",
0050 "spark.glm",
0051 "predict",
0052 "summary",
0053 "spark.kmeans",
0054 "fitted",
0055 "spark.mlp",
0056 "spark.naiveBayes",
0057 "spark.survreg",
0058 "spark.lda",
0059 "spark.posterior",
0060 "spark.perplexity",
0061 "spark.isoreg",
0062 "spark.gaussianMixture",
0063 "spark.als",
0064 "spark.kstest",
0065 "spark.logit",
0066 "spark.decisionTree",
0067 "spark.randomForest",
0068 "spark.gbt",
0069 "spark.bisectingKmeans",
0070 "spark.svmLinear",
0071 "spark.fpGrowth",
0072 "spark.freqItemsets",
0073 "spark.associationRules",
0074 "spark.findFrequentSequentialPatterns",
0075 "spark.assignClusters")
0076
0077 # Job group lifecycle management methods
0078 export("setJobGroup",
0079 "clearJobGroup",
0080 "cancelJobGroup",
0081 "setJobDescription",
0082 "setLocalProperty",
0083 "getLocalProperty")
0084
0085 # Export Utility methods
0086 export("setLogLevel")
0087
0088 exportClasses("SparkDataFrame")
0089
0090 exportMethods("arrange",
0091 "as.data.frame",
0092 "attach",
0093 "broadcast",
0094 "cache",
0095 "checkpoint",
0096 "coalesce",
0097 "collect",
0098 "colnames",
0099 "colnames<-",
0100 "coltypes",
0101 "coltypes<-",
0102 "columns",
0103 "count",
0104 "cov",
0105 "corr",
0106 "covar_samp",
0107 "covar_pop",
0108 "createOrReplaceTempView",
0109 "crossJoin",
0110 "crosstab",
0111 "cube",
0112 "dapply",
0113 "dapplyCollect",
0114 "describe",
0115 "dim",
0116 "distinct",
0117 "drop",
0118 "dropDuplicates",
0119 "dropna",
0120 "dtypes",
0121 "except",
0122 "exceptAll",
0123 "explain",
0124 "fillna",
0125 "filter",
0126 "first",
0127 "freqItems",
0128 "gapply",
0129 "gapplyCollect",
0130 "getNumPartitions",
0131 "group_by",
0132 "groupBy",
0133 "head",
0134 "hint",
0135 "insertInto",
0136 "intersect",
0137 "intersectAll",
0138 "isLocal",
0139 "isStreaming",
0140 "join",
0141 "limit",
0142 "localCheckpoint",
0143 "merge",
0144 "mutate",
0145 "na.omit",
0146 "names",
0147 "names<-",
0148 "ncol",
0149 "nrow",
0150 "orderBy",
0151 "persist",
0152 "printSchema",
0153 "randomSplit",
0154 "rbind",
0155 "registerTempTable",
0156 "rename",
0157 "repartition",
0158 "repartitionByRange",
0159 "rollup",
0160 "sample",
0161 "sample_frac",
0162 "sampleBy",
0163 "saveAsTable",
0164 "saveDF",
0165 "schema",
0166 "select",
0167 "selectExpr",
0168 "show",
0169 "showDF",
0170 "storageLevel",
0171 "subset",
0172 "summarize",
0173 "summary",
0174 "take",
0175 "toJSON",
0176 "transform",
0177 "union",
0178 "unionAll",
0179 "unionByName",
0180 "unique",
0181 "unpersist",
0182 "where",
0183 "with",
0184 "withColumn",
0185 "withColumnRenamed",
0186 "withWatermark",
0187 "write.df",
0188 "write.jdbc",
0189 "write.json",
0190 "write.orc",
0191 "write.parquet",
0192 "write.stream",
0193 "write.text",
0194 "write.ml")
0195
0196 exportClasses("Column")
0197
0198 exportMethods("%<=>%",
0199 "%in%",
0200 "abs",
0201 "acos",
0202 "add_months",
0203 "alias",
0204 "approx_count_distinct",
0205 "approxCountDistinct",
0206 "approxQuantile",
0207 "array_contains",
0208 "array_distinct",
0209 "array_except",
0210 "array_intersect",
0211 "array_join",
0212 "array_max",
0213 "array_min",
0214 "array_position",
0215 "array_remove",
0216 "array_repeat",
0217 "array_sort",
0218 "arrays_overlap",
0219 "array_union",
0220 "arrays_zip",
0221 "asc",
0222 "ascii",
0223 "asin",
0224 "atan",
0225 "atan2",
0226 "avg",
0227 "base64",
0228 "between",
0229 "bin",
0230 "bitwiseNOT",
0231 "bround",
0232 "cast",
0233 "cbrt",
0234 "ceil",
0235 "ceiling",
0236 "collect_list",
0237 "collect_set",
0238 "column",
0239 "concat",
0240 "concat_ws",
0241 "contains",
0242 "conv",
0243 "cos",
0244 "cosh",
0245 "count",
0246 "countDistinct",
0247 "crc32",
0248 "create_array",
0249 "create_map",
0250 "current_date",
0251 "current_timestamp",
0252 "hash",
0253 "cume_dist",
0254 "date_add",
0255 "date_format",
0256 "date_sub",
0257 "date_trunc",
0258 "datediff",
0259 "dayofmonth",
0260 "dayofweek",
0261 "dayofyear",
0262 "decode",
0263 "degrees",
0264 "dense_rank",
0265 "desc",
0266 "element_at",
0267 "encode",
0268 "endsWith",
0269 "exp",
0270 "explode",
0271 "explode_outer",
0272 "expm1",
0273 "expr",
0274 "factorial",
0275 "first",
0276 "flatten",
0277 "floor",
0278 "format_number",
0279 "format_string",
0280 "from_csv",
0281 "from_json",
0282 "from_unixtime",
0283 "from_utc_timestamp",
0284 "getField",
0285 "getItem",
0286 "greatest",
0287 "grouping_bit",
0288 "grouping_id",
0289 "hex",
0290 "histogram",
0291 "hour",
0292 "hypot",
0293 "ifelse",
0294 "initcap",
0295 "input_file_name",
0296 "instr",
0297 "isNaN",
0298 "isNotNull",
0299 "isNull",
0300 "is.nan",
0301 "isnan",
0302 "kurtosis",
0303 "lag",
0304 "last",
0305 "last_day",
0306 "lead",
0307 "least",
0308 "length",
0309 "levenshtein",
0310 "like",
0311 "lit",
0312 "locate",
0313 "log",
0314 "log10",
0315 "log1p",
0316 "log2",
0317 "lower",
0318 "lpad",
0319 "ltrim",
0320 "map_concat",
0321 "map_entries",
0322 "map_from_arrays",
0323 "map_from_entries",
0324 "map_keys",
0325 "map_values",
0326 "max",
0327 "md5",
0328 "mean",
0329 "min",
0330 "minute",
0331 "monotonically_increasing_id",
0332 "month",
0333 "months_between",
0334 "n",
0335 "n_distinct",
0336 "nanvl",
0337 "negate",
0338 "next_day",
0339 "not",
0340 "ntile",
0341 "otherwise",
0342 "over",
0343 "overlay",
0344 "percent_rank",
0345 "pmod",
0346 "posexplode",
0347 "posexplode_outer",
0348 "quarter",
0349 "radians",
0350 "rand",
0351 "randn",
0352 "rank",
0353 "regexp_extract",
0354 "regexp_replace",
0355 "repeat_string",
0356 "reverse",
0357 "rint",
0358 "rlike",
0359 "round",
0360 "row_number",
0361 "rpad",
0362 "rtrim",
0363 "schema_of_csv",
0364 "schema_of_json",
0365 "second",
0366 "sha1",
0367 "sha2",
0368 "shiftLeft",
0369 "shiftRight",
0370 "shiftRightUnsigned",
0371 "shuffle",
0372 "sd",
0373 "sign",
0374 "signum",
0375 "sin",
0376 "sinh",
0377 "size",
0378 "skewness",
0379 "slice",
0380 "sort_array",
0381 "soundex",
0382 "spark_partition_id",
0383 "split_string",
0384 "stddev",
0385 "stddev_pop",
0386 "stddev_samp",
0387 "struct",
0388 "sqrt",
0389 "startsWith",
0390 "substr",
0391 "substring_index",
0392 "sum",
0393 "sumDistinct",
0394 "tan",
0395 "tanh",
0396 "toDegrees",
0397 "toRadians",
0398 "to_csv",
0399 "to_date",
0400 "to_json",
0401 "to_timestamp",
0402 "to_utc_timestamp",
0403 "translate",
0404 "trim",
0405 "trunc",
0406 "unbase64",
0407 "unhex",
0408 "unix_timestamp",
0409 "upper",
0410 "var",
0411 "variance",
0412 "var_pop",
0413 "var_samp",
0414 "weekofyear",
0415 "when",
0416 "window",
0417 "xxhash64",
0418 "year")
0419
0420 exportClasses("GroupedData")
0421 exportMethods("agg")
0422 exportMethods("pivot")
0423
0424 export("as.DataFrame",
0425 "cacheTable",
0426 "clearCache",
0427 "createDataFrame",
0428 "createExternalTable",
0429 "createTable",
0430 "currentDatabase",
0431 "dropTempTable",
0432 "dropTempView",
0433 "listColumns",
0434 "listDatabases",
0435 "listFunctions",
0436 "listTables",
0437 "loadDF",
0438 "read.df",
0439 "read.jdbc",
0440 "read.json",
0441 "read.orc",
0442 "read.parquet",
0443 "read.stream",
0444 "read.text",
0445 "recoverPartitions",
0446 "refreshByPath",
0447 "refreshTable",
0448 "setCheckpointDir",
0449 "setCurrentDatabase",
0450 "spark.lapply",
0451 "spark.addFile",
0452 "spark.getSparkFilesRootDirectory",
0453 "spark.getSparkFiles",
0454 "sql",
0455 "str",
0456 "tableToDF",
0457 "tableNames",
0458 "tables",
0459 "uncacheTable",
0460 "print.summary.GeneralizedLinearRegressionModel",
0461 "read.ml",
0462 "print.summary.KSTest",
0463 "print.summary.DecisionTreeRegressionModel",
0464 "print.summary.DecisionTreeClassificationModel",
0465 "print.summary.RandomForestRegressionModel",
0466 "print.summary.RandomForestClassificationModel",
0467 "print.summary.GBTRegressionModel",
0468 "print.summary.GBTClassificationModel")
0469
0470 export("structField",
0471 "structField.jobj",
0472 "structField.character",
0473 "print.structField",
0474 "structType",
0475 "structType.character",
0476 "structType.jobj",
0477 "structType.structField",
0478 "print.structType")
0479
0480 exportClasses("WindowSpec")
0481
0482 export("partitionBy",
0483 "rowsBetween",
0484 "rangeBetween")
0485
0486 export("windowPartitionBy",
0487 "windowOrderBy")
0488
0489 exportClasses("StreamingQuery")
0490
0491 export("awaitTermination",
0492 "isActive",
0493 "lastProgress",
0494 "queryName",
0495 "status",
0496 "stopQuery")
0497
0498
0499 S3method(print, jobj)
0500 S3method(print, structField)
0501 S3method(print, structType)
0502 S3method(print, summary.GeneralizedLinearRegressionModel)
0503 S3method(print, summary.KSTest)
0504 S3method(print, summary.DecisionTreeRegressionModel)
0505 S3method(print, summary.DecisionTreeClassificationModel)
0506 S3method(print, summary.RandomForestRegressionModel)
0507 S3method(print, summary.RandomForestClassificationModel)
0508 S3method(print, summary.GBTRegressionModel)
0509 S3method(print, summary.GBTClassificationModel)
0510 S3method(structField, character)
0511 S3method(structField, jobj)
0512 S3method(structType, character)
0513 S3method(structType, jobj)
0514 S3method(structType, structField)
