Thank you for the flatten function, it has a bigger functionality than what I need for my project but the examples (which were really, really useful) helped me find a solution.
Instead of accessing the confidence and entity attributes (metadata.confidence and metadata.entity) I was accessing by metadata.value, which instead of returning an error gave null values. In addition, the confidence value (a number) has StringType so before calculating the average I had to convert it to DoubleType. The output generated now is the same as with Spark but using only around 8.5GB of RAM so there's no longer a memory error! Thank you again for all your help! ---- Activado Wed, 20 Apr 2022 14:24:46 +0200 Bjørn Jørgensen <bjornjorgen...@gmail.com> escribió ---- Glad to hear that it works :) Your dataframe is nested with both map, array and struct. I`m using this function to flatten a nested dataframe to rows and columns. from pyspark.sql.types import * from pyspark.sql.functions import * def flatten_test(df, sep="_"): """Returns a flattened dataframe. .. versionadded:: x.X.X Parameters ---------- sep : str Delimiter for flatted columns. Default `_` Notes ----- Don`t use `.` as `sep` It won't work on nested data frames with more than one level. And you will have to use `http://columns.name`. Flattening Map Types will have to find every key in the column. This can be slow. Examples -------- data_mixed = [ { "state": "Florida", "shortname": "FL", "info": {"governor": "Rick Scott"}, "counties": [ {"name": "Dade", "population": 12345}, {"name": "Broward", "population": 40000}, {"name": "Palm Beach", "population": 60000}, ], }, { "state": "Ohio", "shortname": "OH", "info": {"governor": "John Kasich"}, "counties": [ {"name": "Summit", "population": 1234}, {"name": "Cuyahoga", "population": 1337}, ], }, ] data_mixed = spark.createDataFrame(data=data_mixed) data_mixed.printSchema() root |-- counties: array (nullable = true) | |-- element: map (containsNull = true) | | |-- key: string | | |-- value: string (valueContainsNull = true) |-- info: map (nullable = true) | |-- key: string | |-- value: string (valueContainsNull = true) |-- shortname: string (nullable = true) |-- state: string (nullable = true) data_mixed_flat = flatten_test(df, sep=":") data_mixed_flat.printSchema() root |-- shortname: string (nullable = true) |-- state: string (nullable = true) |-- counties:name: string (nullable = true) |-- counties:population: string (nullable = true) |-- info:governor: string (nullable = true) data = [ { "id": 1, "name": "Cole Volk", "fitness": {"height": 130, "weight": 60}, }, {"name": "Mark Reg", "fitness": {"height": 130, "weight": 60}}, { "id": 2, "name": "Faye Raker", "fitness": {"height": 130, "weight": 60}, }, ] df = spark.createDataFrame(data=data) df.printSchema() root |-- fitness: map (nullable = true) | |-- key: string | |-- value: long (valueContainsNull = true) |-- id: long (nullable = true) |-- name: string (nullable = true) df_flat = flatten_test(df, sep=":") df_flat.printSchema() root |-- id: long (nullable = true) |-- name: string (nullable = true) |-- fitness:height: long (nullable = true) |-- fitness:weight: long (nullable = true) data_struct = [ (("James",None,"Smith"),"OH","M"), (("Anna","Rose",""),"NY","F"), (("Julia","","Williams"),"OH","F"), (("Maria","Anne","Jones"),"NY","M"), (("Jen","Mary","Brown"),"NY","M"), (("Mike","Mary","Williams"),"OH","M") ] schema = StructType([ StructField('name', StructType([ StructField('firstname', StringType(), True), StructField('middlename', StringType(), True), StructField('lastname', StringType(), True) ])), StructField('state', StringType(), True), StructField('gender', StringType(), True) ]) df_struct = spark.createDataFrame(data = data_struct, schema = schema) df_struct.printSchema() root |-- name: struct (nullable = true) | |-- firstname: string (nullable = true) | |-- middlename: string (nullable = true) | |-- lastname: string (nullable = true) |-- state: string (nullable = true) |-- gender: string (nullable = true) df_struct_flat = flatten_test(df_struct, sep=":") df_struct_flat.printSchema() root |-- state: string (nullable = true) |-- gender: string (nullable = true) |-- name:firstname: string (nullable = true) |-- name:middlename: string (nullable = true) |-- name:lastname: string (nullable = true) """ # compute Complex Fields (Arrays, Structs and Map Types) in Schema complex_fields = dict([(http://field.name, field.dataType) for field in df.schema.fields if type(field.dataType) == ArrayType or type(field.dataType) == StructType or type(field.dataType) == MapType]) while len(complex_fields) !=0: col_name = list(complex_fields.keys())[0] #print ("Processing :"+col_name+" Type : "+str(type(complex_fields[col_name]))) # if StructType then convert all sub elements to columns. # i.e. flatten structs if (type(complex_fields[col_name]) == StructType): expanded = [col(col_name + '.' + k).alias(col_name + sep + k) for k in [http://n.name for n in complex_fields[col_name]]] df = df.select("*", *expanded).drop(col_name) # if ArrayType then add the Array Elements as Rows using the explode function # i.e. explode Arrays elif (type(complex_fields[col_name]) == ArrayType): df = df.withColumn(col_name, explode_outer(col_name)) # if MapType then convert all sub elements to columns. # i.e. flatten elif (type(complex_fields[col_name]) == MapType): keys_df = df.select(explode_outer(map_keys(col(col_name)))).distinct() keys = list(map(lambda row: row[0], keys_df.collect())) key_cols = list(map(lambda f: col(col_name).getItem(f) .alias(str(col_name + sep + f)), keys)) drop_column_list = [col_name] df = df.select([col_name for col_name in df.columns if col_name not in drop_column_list] + key_cols) # recompute remaining Complex Fields in Schema complex_fields = dict([(http://field.name, field.dataType) for field in df.schema.fields if type(field.dataType) == ArrayType or type(field.dataType) == StructType or type(field.dataType) == MapType]) return df Please use "answer to all" on your emails so that the user group can see the emails. :) ons. 20. apr. 2022 kl. 14:05 skrev Xavier Gervilla <mailto:xavier.gervi...@datapta.com>: Changing this worked! spark = sparknlp.start(spark32=True) I'm adapting the rest of the code now, understanding the new schema and debugging and I've found this https://github.com/JohnSnowLabs/spark-nlp-workshop/blob/master/tutorials/Certification_Trainings/Public/3.SparkNLP_Pretrained_Models.ipynb with many examples that has helped get a clearer idea of how to implement the pipeline. This is the schema resulting after applying the pipeline and selecting only the sentiment and the entities obtained: |-- sentiment: array (nullable = true) | |-- element: struct (containsNull = true) | | |-- annotatorType: string (nullable = true) | | |-- begin: integer (nullable = false) | | |-- end: integer (nullable = false) | | |-- result: string (nullable = true) | | |-- metadata: map (nullable = true) | | | |-- key: string | | | |-- value: string (valueContainsNull = true) | | |-- embeddings: array (nullable = true) | | | |-- element: float (containsNull = false) |-- entities: array (nullable = true) | |-- element: struct (containsNull = true) | | |-- annotatorType: string (nullable = true) | | |-- begin: integer (nullable = false) | | |-- end: integer (nullable = false) | | |-- result: string (nullable = true) | | |-- metadata: map (nullable = true) | | | |-- key: string | | | |-- value: string (valueContainsNull = true) | | |-- embeddings: array (nullable = true) | | | |-- element: float (containsNull = false) I have used explode on both to change from an array per row to a single sentiment per row, which works fine, but I'm having trouble handling null/None values, specifically both with result and metadata. On sentiment it's the equivalent of 'negative'/'positive'/'na' and {confidence -> value} for example whereas on entities it's 'entity predicted' and {entity -> label, sentence -> value, chunk -> value}, with label being MISC, ORG... I've tried using drop like this df.drop(col('nerEx.result') == None) but I still get null/None values. Thank you again for your help, this is more related to specific debugging but your suggestion of using SparkNLP was very useful. ---- Activado Tue, 19 Apr 2022 21:41:26 +0200 Bjørn Jørgensen <mailto:bjornjorgen...@gmail.com> escribió ---- https://github.com/JohnSnowLabs/spark-nlp#packages-cheatsheet change spark = sparknlp.start() to spark = sparknlp.start(spark32=True) tir. 19. apr. 2022 kl. 21:10 skrev Bjørn Jørgensen <mailto:bjornjorgen...@gmail.com>: Yes, there are some that have that issue. Please open a new issue at https://github.com/JohnSnowLabs/spark-nlp/issues and they will help you. tir. 19. apr. 2022 kl. 20:33 skrev Xavier Gervilla <mailto:xavier.gervi...@datapta.com>: Thank you for your advice, I had small knowledge of Spark NLP and I thought it was only possible to use with models that required training and therefore my project wasn’t the case. I'm trying now to build the project again with SparkNLP but when I try to load a pretrained model from JohnSnowLabs I get an error (py4j.protocol.Py4JJavaError: An error occurred while calling z:com.johnsnowlabs.nlp.pretrained.PythonResourceDownloader.getDownloadSize.). This is the new basic code to develop the project again: spark = sparknlp.start() pipelineName = 'analyze_sentiment' pipeline = PretrainedPipeline(pipelineName, 'en') #this is the line that generates the error rawTweets = spark.readStream.format('socket').option('host', 'localhost').option('port',9008).load() allTweets = rawTweets.selectExpr('CAST(value AS STRING)').withColumnRenamed('value', 'text').dropDuplicates('text') sentPred = pipeline.transform(allTweets) query = sentPred.writeStream.outputMode('complete').format('console').start() query.awaitTermination() Spark version is 3.2.1 and SparkNLP version is 3.4.3, while Java version is 8. I've tried with a different model but the error is still the same, so what could be causing it? If this error is solved I think SparkNLP will be the solution I was looking for to reduce memory consumption so thank you again for suggesting it. El 18 abr 2022, a las 21:07, Bjørn Jørgensen <mailto:bjornjorgen...@gmail.com> escribió: When did SpaCy have support for Spark? Try https://nlp.johnsnowlabs.com/ it`s made for spark. They have a lot of notebooks at https://github.com/JohnSnowLabs/spark-nlp and they public user guides at https://towardsdatascience.com/introduction-to-spark-nlp-foundations-and-basic-components-part-i-c83b7629ed59 man. 18. apr. 2022 kl. 16:17 skrev Sean Owen <mailto:sro...@gmail.com>: It looks good, are you sure it even starts? the problem I see is that you send a copy of the model from the driver for every task. Try broadcasting the model instead. I'm not sure if that resolves it but would be a good practice. On Mon, Apr 18, 2022 at 9:10 AM Xavier Gervilla <mailto:xavier.gervi...@datapta.com> wrote: Hi Team,https://stackoverflow.com/questions/71841814/is-there-a-way-to-prevent-excessive-ram-consumption-with-the-spark-configuration I'm developing a project that retrieves tweets on a 'host' app, streams them to Spark and with different operations with DataFrames obtains the Sentiment of the tweets and their entities applying a Sentiment model and a NER model respectively. The problem I've come across is that when applying the NER model, the RAM consumption increases until the program stops with a memory error because there's no memory left to execute. In addition, on SparkUI I've seen that there's only one executor running, the executor driver, but using htop on the terminal I see that the 8 cores of the instance are executing at 100%. The SparkSession is only configured to receive the tweets from the socket that connects with the second program that sends the tweets. The DataFrame goes through some processing to obtain other properties of the tweet like its sentiment (which causes no error even with less than 8GB of RAM) and then the NER is applied. spark = SparkSession.builder.appName("TwitterStreamApp").getOrCreate() rawTweets = spark.readStream.format("socket").option("host", "localhost").option("port",9008).load() tweets = rawTweets.selectExpr("CAST(value AS STRING)") #prior processing of the tweets sentDF = other_processing(tweets) #obtaining the column that contains the list of entities from a tweet nerDF = ner_classification(sentDF) This is the code of the functions related to obtaining the NER, the "main call" and the UDF function. nerModel = spacy.load("en_core_web_sm") #main call, applies the UDF function to every tweet from the "tweet" column def ner_classification(words): ner_list = udf(obtain_ner_udf, ArrayType(StringType())) words = words.withColumn("nerlist", ner_list("tweet")) return words #udf function def obtain_ner_udf(words): #if the tweet is empty return None if words == "": return None #else: applying the NER model (Spacy en_core_web_sm) entities = nerModel(words) #returns a list of the form ['entity1_label1', 'entity2_label2',...] return [ word.text + '_' + word.label_ for word in entities.ents ] And lastly I map each entity with the sentiment from its tweet and obtain the average sentiment of the entity and the number of appearances. flattenedNER = nerDF.select(nerDF.sentiment, explode(nerDF.nerlist)) flattenedNER.registerTempTable("df") querySelect = "SELECT col as entity, avg(sentiment) as sentiment, count(col) as count FROM df GROUP BY col" finalDF = spark.sql(querySelect) query = finalDF.writeStream.foreachBatch(processBatch).outputMode("complete").start() The resulting DF is processed with a function that separates each column in a list and prints it. def processBatch(df, epoch_id): entities = [str(t.entity) for t in df.select("entity").collect()] sentiments = [float(t.sentiment) for t in df.select("sentiment").collect()] counts = [int(row.asDict()['count']) for row in df.select("count").collect()] print(entities, sentiments, counts) At first I tried with other NER models from Flair they have the same effect, after printing the first batch memory use starts increasing until it fails and stops the execution because of the memory error. When applying a "simple" function instead of the NER model, such as return words.split() on the UDF there's no such error so the data ingested should not be what's causing the overload but the model. Is there a way to prevent the excessive RAM consumption? Why is there only the driver executor and no other executors are generated? How could I prevent it from collapsing when applying the NER model? Thanks in advance! -- Bjørn Jørgensen Vestre Aspehaug 4, 6010 Ålesund Norge +47 480 94 297 -- Bjørn Jørgensen Vestre Aspehaug 4, 6010 Ålesund Norge +47 480 94 297 -- Bjørn Jørgensen Vestre Aspehaug 4, 6010 Ålesund Norge +47 480 94 297 -- Bjørn Jørgensen Vestre Aspehaug 4, 6010 Ålesund Norge +47 480 94 297
