centrifuge

KO | EN

gitlite — search

#javascript #ai #go #rust #deep-learning #linux #java #windows #awesome #kubernetes #vue #awesome-list #security #mcp

(★ 19)

Data quality tools for Big Data

centrifuge Latest Version Download

Download Latest Version (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

[![Build Status](https://travis-ci.org/UNIVALENCE/centrifuge.png?branch=master)](https://travis-ci.org/UNIVALENCE/centrifuge)
[![codecov](https://codecov.io/gh/UNIVALENCE/centrifuge/branch/master/graph/badge.svg)](https://codecov.io/gh/UNIVALENCE/centrifuge)

# centrifuge
Data quality tools for Big Data

* Scala 2.11.7
* JDK 8
* Spark 2.1.1

See this [video](https://www.youtube.com/watch?v=t24sUF2zWLY) for in-depth coverage of the project

### Quick start
centrifuge will be published to Maven Central soon, so when it is done you will be able to get centrifuge with adding this to your `build.sbt`:
    
```scala    
val sparkV = "2.1.1"
    
libraryDependencies ++= Seq(
  "io.univalence" %% "centrifuge" % "0.1"
)
```   
 
##### Using Annotations
We import relevant libraries and define 3 case classes

```scala
package io.univalence.centrifuge.sql

import io.univalence.centrifuge.{Annotation, AnnotationSql, Result}
import org.apache.spark.sql.SparkSession
import io.univalence.centrifuge.implicits._
import ss.implicits._

case class Person(name: String, age: Int)
case class PersonWithAnnotations(name: String, age: Int, annotations: Seq[Annotation])
case class BetterPerson(name:String,age:Option[Int])
```

Here we create 3 instances of Person

```scala
val joe = Person("Joe", 14)
val timmy = Person("Timmy",8)
val invalid = Person("",-1)    
```

Making a Dataset from the 3 Person we just created

```scala
val ds = ss.sparkContext.makeRDD(Seq(joe,timmy,invalid)).toDS()
ds.createOrReplaceTempView("personRaw")
```

bla

```scala    
ss.registerTransformation[Int,Int]("checkAge", {
  case i if i < 0 => Result.fromError("INVALID_AGE")
  case i if i < 13 => Result.fromWarning(i, "UNDER_13")
  case i if i > 140 => Result.fromError("OVER_140")
  case i => Result.pure(i)
})
``` 
  
Basic select

```scala    
ss.sql("select name, checkAge(age) as age, age as ageRaw from personRaw").show(false)
```

Result

+-----+----+------+
    |name |age |ageRaw|
    +-----+----+------+
    |Joe  |14  |14    |
    |Timmy|8   |8     |
    |     |null|-1    |
    +-----+----+------+

This time using BetterPerson (a Person with Option)

```scala
ss.sql("select name, checkAge(age) as age from personRaw").as[BetterPerson].collect().foreach(println)
```

As expected
    
    BetterPerson(Joe,Some(14))
    BetterPerson(Timmy,Some(8))
    BetterPerson(,None)
    
Now we include Annotations

```scala
ss.sql("select name, checkAge(age) as age from personRaw").includeAnnotations.show(false)
```
    
Result

+-----+----+--------------------------------------------+
    |name |age |annotations                                 |
    +-----+----+--------------------------------------------+
    |Joe  |14  |[]                                          |
    |Timmy|8   |[[UNDER_13,age,WrappedArray(age),false,1]]  |
    |     |null|[[INVALID_AGE,age,WrappedArray(age),true,1]]|
    +-----+----+--------------------------------------------+

We now have annotations for each Person in the DataSet.

Let's take a look at what an Annotation is:

```scala
Annotation(message: String,
           onField: Option[String] = None,
           fromFields: Vector[String] = Vector.empty,
           isError: Boolean,
           count: Long)
```

* **message** contains a relevant information
* **onField** is the name of the field the message is from
* **fromFields** is
* **isError** is **true** if the data is not within acceptable values else it's **false**
* **count** is the number of time the value has appeared for the same field

### TODO / Roadmap
Integrate a reworked version of Excelsius in order to have a better DataFrame.show()

To go from

++-----+----+------------------------------------------------+
    +|name |age |annotations                                     |
    +|     |    |message     |onField |fromFields |isError |count|
    ++-----+----+------------+--------+-----------+--------+-----+
    +|Joe  |14  |                  ---                           |
    +|Timmy|8   |UNDER_13    |age     |age        |false   |1    |
    +|     |null|INVALID_AGE |age     |age        |true    |1    |
    +|  ------  |EMPTY_STRING|name    |name1      |true    |1    |
    +|  ------  |     ---------       |name2      |    ----      |
    ++-----+----+------------+--------+-----------+--------+-----+

"-" is a placeholder and a better symbol/ascii art will be used
## License
centrifuge is licensed under the Apache License, Version 2.0 (the “License”); you may not use this software except in compliance with the License.

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an “AS IS” BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

centrifuge

centrifuge Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

centrifuge

centrifuge Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

Download README