Data preparation

Visual data preparation in DSS lets you create data cleansing, normalization and enrichment scripts in a visual and interactive way.

You can create these scripts directly in a Prepare recipe, or in a Visual Analysis that can be deployed to the Flow as a Prepare recipe.

Note

For a step by step introduction to the data preparation component of Data Science Studio, we recommend that you follow the:

• Quick Start | Dataiku for data preparation

• Visual recipes course on the Dataiku Academy.

• How to Copy Prepare Recipe Steps
  • Copying Steps
• Sampling
• Execution engines
  • Design of the preparation
  • Execution in analysis
  • Execution of the recipe
  • Details on the in-database (SQL) engine
  • Details on the Spark engine
• Processors reference
  • Extract from array
  • Fold an array
  • Sort array
  • Concatenate JSON arrays
  • Discretize (bin) numerical values
  • Change coordinates system
  • Copy column
  • Rename columns
  • Concatenate columns
  • Delete/Keep columns by name
  • Column Pseudonymization
  • Count occurrences
  • Convert currencies
  • Split currencies in column
  • Create if, then, else statements
  • Extract date elements
  • Compute difference between dates
  • Format date with custom format
  • Parse to standard date format
  • Split e-mail addresses
  • Enrich from French department
  • Enrich from French postcode
  • Enrich with build context
  • Enrich with record context
  • Extract ngrams
  • Extract numbers
  • Fill column
  • Fill empty cells with fixed value
  • Impute with computed value
  • Filter rows/cells on date
  • Filter rows/cells with formula
  • Filter invalid rows/cells
  • Filter rows/cells on numerical range
  • Filter rows/cells on value
  • Find and replace
  • Flag rows/cells on date range
  • Flag rows with formula
  • Flag invalid rows
  • Flag rows on numerical range
  • Flag rows on value
  • Fold multiple columns
  • Fold multiple columns by pattern
  • Fold object keys
  • Formula
  • Fuzzy join with other dataset (memory-based)
  • Generate Big Data
  • Compute distance between geospatial objects
  • Extract from geo column
  • Geo-join
  • Resolve GeoIP
  • Create area around a geopoint
  • Create GeoPoint from lat/lon
  • Extract lat/lon from GeoPoint
  • Extract with grok
  • Flag holidays
  • Split invalid cells into another column
  • Join with other dataset (memory-based)
  • Extract with JSONPath
  • Group long-tail values
  • Compute the average of numerical values
  • Translate values using meaning
  • Normalize measure
  • Merge long-tail values
  • Move columns
  • Negate boolean value
  • Force numerical range
  • Generate numerical combinations
  • Convert number formats
  • Nest columns
  • Unnest object (flatten JSON)
  • Extract with regular expression
  • Pivot
  • Python function
  • Split HTTP Query String
  • Remove rows where cell is empty
  • Round numbers
  • Simplify text
  • Split and fold
  • Split into chunks
  • Split and unfold
  • Split column
  • Switch case
  • Transform string
  • Tokenize text
  • Transpose rows to columns
  • Triggered unfold
  • Unfold
  • Unfold an array
  • Convert a UNIX timestamp to a date
  • Fill empty cells with previous/next value
  • Split URL (into protocol, host, port, …)
  • Classify User-Agent
  • Generate a best-effort visitor id
  • Zip JSON arrays
• Filtering and flagging rows
  • Common filtering actions
  • Columns selection
  • Filter on value
  • Filter on numerical range
  • Filter on date range
  • Filter on formula
  • Filter on bad meaning
• Managing dates
  • Working with dates in Preparation
  • Parsing Dates
  • Using dates
• Reshaping
  • Split and Fold
  • Fold multiple columns
  • Fold multiple columns by pattern
  • Unfold
  • Unfold an array
  • Split and Unfold
  • Triggered Unfold
• Geographic processors
  • Geopoint converters
  • Resolve GeoIP
  • Reverse geocoding
  • Zipcode geocoding
  • Change coordinates system
  • Compute distances between geospatial objects
  • Create area around a geopoint
  • Extract from geo column