Call Girls Indiranagar Just Call 👗 7737669865 👗 Top Class Call Girl Service B...
U-SQL Killer Scenarios: Taming the Data Science Monster with U-SQL and Big Cognition (SQL Konferenz 2017)
1. U-SQL Killer Scenarios:
Taming the Data
Science Monster with U-
SQL and Big Cognition
Michael Rys
Principal Program Manager, Big Data
Microsoft
@MikeDoesBigData, usql@microsoft.com
2. Agenda • Introduction to U-SQL Extensibility
• U-SQL Cognitive Services
• More Custom Image processing
• Python in U-SQL
• R in U-SQL
• JSON processing
4. What are UDOs?
• User-Defined Extractors
• User-Defined Outputters
• User-Defined Processors
• Take one row and produce one row
• Pass-through versus transforming
• User-Defined Appliers
• Take one row and produce 0 to n rows
• Used with OUTER/CROSS APPLY
• User-Defined Combiners
• Combines rowsets (like a user-defined join)
• User-Defined Reducers
• Take n rows and produce m rows (normally m<n)
• Scaled out with explicit U-SQL Syntax that takes a
UDO instance (created as part of the execution):
• EXTRACT
• OUTPUT
• CROSS APPLY
Custom Operator Extensions
Scaled out by U-SQL
• PROCESS
• COMBINE
• REDUCE
5. [SqlUserDefinedExtractor]
public class DriverExtractor : IExtractor
{
private byte[] _row_delim;
private string _col_delim;
private Encoding _encoding;
// Define a non-default constructor since I want to pass in my own parameters
public DriverExtractor( string row_delim = "rn", string col_delim = ",“
, Encoding encoding = null )
{
_encoding = encoding == null ? Encoding.UTF8 : encoding;
_row_delim = _encoding.GetBytes(row_delim);
_col_delim = col_delim;
} // DriverExtractor
// Converting text to target schema
private void OutputValueAtCol_I(string c, int i, IUpdatableRow outputrow)
{
var schema = outputrow.Schema;
if (schema[i].Type == typeof(int))
{
var tmp = Convert.ToInt32(c);
outputrow.Set(i, tmp);
}
...
} //SerializeCol
public override IEnumerable<IRow> Extract( IUnstructuredReader input
, IUpdatableRow outputrow)
{
foreach (var row in input.Split(_row_delim))
{
using(var s = new StreamReader(row, _encoding))
{
int i = 0;
foreach (var c in s.ReadToEnd().Split(new[] { _col_delim }, StringSplitOptions.None))
{
OutputValueAtCol_I(c, i++, outputrow);
} // foreach
} // using
yield return outputrow.AsReadOnly();
} // foreach
} // Extract
} // class DriverExtractor
UDO model
Marking UDOs
Parameterizing UDOs
UDO signature
UDO-specific processing
pattern
Rowsets and their schemas
in UDOs
Setting results
• By position
• By name
6. • .Net API provided to build UDOs
• Any .Net language usable
• however only C# is first-class in tooling
• Use U-SQL specific .Net DLLs
• Deploying UDOs
• Compile DLL
• Upload DLL to ADLS
• register with U-SQL script
• VisualStudio provides tool support
• UDOs can
• Invoke managed code
• Invoke native code deployed with UDO assemblies
• Invoke other language runtimes (e.g., Python, R)
• be scaled out by U-SQL execution framework
• UDOs cannot
• Communicate between different UDO invocations
• Call Webservices/Reach outside the vertex boundary
How to specify
UDOs?
9. Managing
Assemblies
Create assemblies
Reference assemblies
Enumerate assemblies
Drop assemblies
VisualStudio makes registration easy!
• CREATE ASSEMBLY db.assembly FROM @path;
• CREATE ASSEMBLY db.assembly FROM byte[];
• Can also include additional resource files
• REFERENCE ASSEMBLY db.assembly;
• Referencing .Net Framework Assemblies
• Always accessible system namespaces:
• U-SQL specific (e.g., for SQL.MAP)
• All provided by system.dll system.core.dll
system.data.dll, System.Runtime.Serialization.dll,
mscorelib.dll (e.g., System.Text,
System.Text.RegularExpressions, System.Linq)
• Add all other .Net Framework Assemblies with:
REFERENCE SYSTEM ASSEMBLY [System.XML];
• Enumerating Assemblies
• Powershell command
• U-SQL Studio Server Explorer and Azure Portal
• DROP ASSEMBLY db.assembly;
10. USING clause
'USING' csharp_namespace
| Alias '=' csharp_namespace_or_class.
Examples:
DECLARE @ input string = "somejsonfile.json";
REFERENCE ASSEMBLY [Newtonsoft.Json];
REFERENCE ASSEMBLY [Microsoft.Analytics.Samples.Formats];
USING Microsoft.Analytics.Samples.Formats.Json;
@data0 = EXTRACT IPAddresses string
FROM @input
USING new JsonExtractor("Devices[*]");
USING json =
[Microsoft.Analytics.Samples.Formats.Json.JsonExtractor];
@data1 = EXTRACT IPAddresses string
FROM @input
USING new json("Devices[*]");
11. DEPLOY
RESOURCE
Syntax:
'DEPLOY' 'RESOURCE' file_path_URI { ',' file_path_URI }.
Example:
DEPLOY RESOURCE "/config/configfile.xml", "package.zip";
Semantics:
• Files have to be in ADLS or WASB
• Files are deployed to vertex and are accessible from any custom
code
Limits:
• Single resource file limit is 400MB
• Overall limit for deployed resource files is 3GB
14. REFERENCE ASSEMBLY ImageCommon;
REFERENCE ASSEMBLY FaceSdk;
REFERENCE ASSEMBLY ImageEmotion;
REFERENCE ASSEMBLY ImageTagging;
REFERENCE ASSEMBLY ImageOcr;
@imgs =
EXTRACT FileName string, ImgData byte[]
FROM @"/images/{FileName:*}.jpg"
USING new Cognition.Vision.ImageExtractor();
// Extract the number of objects on each image and tag them
@objects =
PROCESS @imgs
PRODUCE FileName,
NumObjects int,
Tags string
READONLY FileName
USING new Cognition.Vision.ImageTagger();
OUTPUT @objects
TO "/objects.tsv"
USING Outputters.Tsv();
Imaging
15. REFERENCE ASSEMBLY [TextCommon];
REFERENCE ASSEMBLY [TextSentiment];
REFERENCE ASSEMBLY [TextKeyPhrase];
@WarAndPeace =
EXTRACT No int,
Year string,
Book string, Chapter string,
Text string
FROM @"/usqlext/samples/cognition/war_and_peace.csv"
USING Extractors.Csv();
@sentiment =
PROCESS @WarAndPeace
PRODUCE No,
Year,
Book, Chapter,
Text,
Sentiment string,
Conf double
USING new Cognition.Text.SentimentAnalyzer(true);
OUTPUT @sentinment
TO "/sentiment.tsv"
USING Outputters.Tsv();
Text Analysis
16. U-SQL/Cognitive
Example
• Identify objects in images (tags)
• Identify faces and emotions and images
• Join datasets – find out which tags are associated with happiness
REFERENCE ASSEMBLY ImageCommon;
REFERENCE ASSEMBLY FaceSdk;
REFERENCE ASSEMBLY ImageEmotion;
REFERENCE ASSEMBLY ImageTagging;
@objects =
PROCESS MegaFaceView
PRODUCE FileName, NumObjects int, Tags string
READONLY FileName
USING new Cognition.Vision.ImageTagger();
@tags =
SELECT FileName, T.Tag
FROM @objects
CROSS APPLY
EXPLODE(SqlArray.Create(Tags.Split(';')))
AS T(Tag)
WHERE T.Tag.ToString().Contains("dog") OR
T.Tag.ToString().Contains("cat");
@emotion_raw =
PROCESS MegaFaceView
PRODUCE FileName string, NumFaces int, Emotion string
READONLY FileName
USING new Cognition.Vision.EmotionAnalyzer();
@emotion =
SELECT FileName, T.Emotion
FROM @emotion_raw
CROSS APPLY
EXPLODE(SqlArray.Create(Emotion.Split(';')))
AS T(Emotion);
@correlation =
SELECT T.FileName, Emotion, Tag
FROM @emotion AS E
INNER JOIN
@tags AS T
ON E.FileName == T.FileName;
Images
Objects Emotions
filter
join
aggregat
e
17. Python Processing
Python
Author Tweet
MikeDoesBigData @AzureDataLake: Come and see the #TR24 sessions on #USQL
AzureDataLake What are your recommendations for #TR24? @MikeDoesBigData
Author Mentions Topics
MikeDoesBigData {@AzureDataLake} {#TR24, #USQL}
AzureDataLake {@MikeDoesBigData} {#TR24}
18. REFERENCE ASSEMBLY [ExtPython];
DECLARE @myScript = @"
def get_mentions(tweet):
return ';'.join( ( w[1:] for w in tweet.split() if w[0]=='@' ) )
def usqlml_main(df):
del df['time']
del df['author']
df['mentions'] = df.tweet.apply(get_mentions)
del df['tweet']
return df
";
@t =
SELECT * FROM
(VALUES
("D1","T1","A1","@foo Hello World @bar"),
("D2","T2","A2","@baz Hello World @beer")
) AS D( date, time, author, tweet );
@m =
REDUCE @t ON date
PRODUCE date string, mentions string
USING new Extension.Python.Reducer(pyScript:@myScript);
Use U-SQL to create a massively
distributed program.
Executing Python code across many
nodes.
Using standard libraries such as
numpy and pandas.
Documentation:
https://docs.microsoft.com/en-
us/azure/data-lake-analytics/data-
lake-analytics-u-sql-python-
extensions
Python
Extensions
23. Image Processing • Image processing assembly
• Uses System.Drawing
• Exposes
• Extractors
• Outputter
• Processor
• User-defined Functions
• Trade-offs
• Column memory limits:
Image Extractor vs Feature
Extractor
• Main memory pressures in vertex:
UDFs vs Processor vs Extractor
24. JSON Processing
How do I extract data from JSON documents?
https://github.com/Azure/usql/tree/master/Examples/DataFormats
https://github.com/Azure/usql/tree/master/Examples/JSONExamples
25. • Architecture of Sample Format Assembly
• Single JSON document per file: Use JsonExtractor
• Multiple JSON documents per file:
• Do not allow row delimiter (e.g., CR/LF) in JSON
• Use built-in Text Extractor to extract
• Use JsonTuple to schematize (with CROSS APPLY)
• Currently loads full JSON document into memory
• better to use JSONReader Processing if docs are large
Microsoft.Analytics.Samples.Formats
NewtonSoft.Json Microsoft.Hadoop.AvroSystem.Xml
JSON
Processin
g
26. JSON
Processin
g
@json =
EXTRACT personid int,
name string,
addresses string
FROM @input
USING new Json.JsonExtractor(“[*].person");
@person =
SELECT personid,
name,
Json.JsonFunctions.JsonTuple(addresses)["address"] AS address_array
FROM @json;
@addresses = SELECT personid, name, Json.JsonFunctions.JsonTuple(address) AS address
FROM @person
CROSS APPLY
EXPLODE (Json.JsonFunctions.JsonTuple(address_array).Values) AS A(address);
@result =
SELECT personid,
name,
address["addressid"]AS addressid,
address["street"]AS street,
address["postcode"]AS postcode,
address["city"]AS city
FROM @addresses;
29. UDO Tips and
Warnings
• Tips when Using UDOs:
• READONLY clause to allow pushing predicates through UDOs
• REQUIRED clause to allow column pruning through UDOs
• PRESORT on REDUCE if you need global order
• Hint Cardinality if it does choose the wrong plan
• Warnings and better alternatives:
• Use SELECT with UDFs instead of PROCESS
• Use User-defined Aggregators instead of REDUCE
• Learn to use Windowing Functions (OVER expression)
• Good use-cases for
PROCESS/REDUCE/COMBINE:
• The logic needs to dynamically access the input and/or output
schema.
E.g., create a JSON doc for the data in the row where the
columns are not known apriori.
• Your UDF based solution creates too much memory pressure and
you can write your code more memory efficient in a UDO
• You need an ordered Aggregator or produce more than 1 row per
group