Tabular transformations are typically used to refine data in a way that makes insights and understanding obvious. They range from simple augmentation of data by deriving new columns from existing ones to summarizing data into different groups. Often transformations are used in conjunction with other tasks, such as visualization and modeling.  The Wolfram Language provides a set of highly optimized tabular transformation functions that are easy to use and scale to large data.

Modifying Tabular Objects

TransformColumns — compute columns from values in each row

ConstructColumns — extract or construct columns

ColumnwiseThread  ▪  ColumnwiseValue  ▪  Elementwise

Reordering Tabular Objects »

SortBy  ▪  ReverseSortBy  ▪  NumericalSort  ▪  AlphabeticSort  ▪  ...

Simple and Grouped Aggregation »

AggregateRows — simple or grouped aggregation

Total  ▪  Count  ▪  Mean  ▪  Median  ▪  Max  ▪  StandardDeviation  ▪  ...

PivotTable — compute a pivot table with multiple row and column pivots and aggregation

IncludeGroupAggregates — whether to include aggregation for each subgroup

Tabular Objects »

Tabular — column-oriented tabular data

ToTabular  ▪  TabularStructure  ▪  Select  ▪  TakeLargestBy  ▪  ...

Tabular Data Cleaning »

TransformMissing — how to handle missing values, impute values, etc.

RenameColumns  ▪  CastColumns  ▪  PivotToColumns  ▪  TransformAnomalies  ▪  ...

Combining Tabular Objects

Join — joining tabular objects by stacking them

JoinAcross — support for relational joins of two tabular objects

LeftJoinAcross — specific support for the left join of two tabular objects

ColumnwiseCombine — combine several tabular objects, comparing values of selected columns