TransformColumns

TransformColumns[tab,ncolf]

adds a new column with name ncol by transforming the tabular data tab using the function f applied to each row.

TransformColumns[tab,{ncol1f1,ncol2f2,…}]

adds several new columns ncoli by successively applying the functions fi to each row.

TransformColumns[transfs]

represents an operator form of TransformColumns that can be applied to tabular data.

Details and Options

  • TransformColumns is typically used to add new columns or modify existing columns, keeping the rest.
  • Possible forms of tabular data tab include:
  • Tabular[…]type-consistent tabular data
    Dataset[…]general hierarchical data
    {assoc1,assoc2,…}list of associations with common keys
    matrixmatrix of values
  • If the column ncoli already exists, it is replaced with the newly created one. Otherwise, a new column is added after the existing ones.
  • TransformColumns[tab,{col,ncolf}] adds a new column named ncol after the existing column col, instead of appending it at the end.
  • For tabular data atab without column keys, TransformColumns[atab,{f1,f2,…,fn}] appends n new anonymous columns by successively applying the functions fi to each row.
  • Functions fi are applied to individual rows of the input tabular data tab, with the row being an association <|col1val1,…|> if tab has column keys or a list {val1,…} if tab does not have column keys.
  • When several functions are used, TransformColumns[tab,{…}] is effectively equivalent to Fold[TransformColumns,tab,{…}], allowing each function to access the columns previously added.
  • The syntax colNothing can be used to remove the existing column named col. »
  • TransformColumns[transfs][tab] is equivalent to TransformColumns[tab,transfs].

Examples

open allclose all

Basic Examples  (4)

Transform the difference of columns "a" and "b" into a new column "c":

Double the column "a" and halve the column "b":

Transform existing columns into a new column:

Transform columns on an out-of-core tabular:

Fetch the population and area of the Central American countries, expressed in Tabular form:

Compute population density for each country:

Scope  (18)

Transformations  (7)

Transform an existing column in a Tabular object:

Split an existing column in a Tabular object:

TransformColumns with a function f acts on whole rows as associations if columns are named:

On a Tabular object with anonymous columns, the rows are expressed as lists:

Use the operator form of TransformColumns:

Use "a"Nothing to remove column "a" from the result:

Remove three columns from the result:

Add the second and fourth columns to make a fifth column in a matrix:

With Apply around the function, there is an argument for each column:

Use Key in the Function body of Tabular transformations when the key is a general expression:

Input Data  (3)

Transform a list of associations:

Modify an existing column:

Create a new column as a function of existing columns:

Transform a Dataset object:

Transform the columns of a normal matrix:

Columnwise Operations  (4)

Use ColumnwiseValue to subtract the mean of a column from its values:

Find which elements in a column are above the median:

Include a constant column with the median value:

Compute the median only once by using the operator form:

Use ColumnwiseThread to compute a vector-valued transformation of entire columns:

With ColumnwiseValue, you get the same list for each row:

Take a Tabular object:

TransformColumns by default operates sequentially on each row:

TransformColumns with ColumnwiseValue[expr] evaluates expr once first, then acts on each row:

TransformColumns with ColumnwiseThread[expr] evaluates expr once, acting on whole columns:

Placement of Columns  (2)

Take a Tabular object:

By default, TransformColumns places new columns at the end:

Use column "b" as an anchor column after which the new column is placed:

Place the new column after column "a":

Place the new column before column "a":

Use column names without a transformation to anchor column placement in the result:

By default, new columns are appended on the right:

Place "e" before "a" and "f" after "b":

Place "e" first and "f" after "b" and remove "a" from the result:

Missing Fallback  (2)

Use MissingFallback to fill in missing values when transforming a Tabular object:

Use row values from the previous column:

Use row values from the next column:

Use a fixed value:

Use functions of columns in MissingFallback to fill in missing values when transforming a Tabular object:

Use the sum of row values from other columns:

Use the mean of row values from neighboring columns:

Applications  (3)

Medical Data  (1)

This data contains information on kidney transplant patients—time in weeks to death or on-study time since transplant at the given age:

Use TransformColumns to give meaning to the numerically encoded data:

Now use the categories to compute mean survival time in years with PivotTable:

Weather Data  (1)

Weather data from JFK airport in C, mbar and km/h:

Compute the windchill factor:

Create a Tabular object with new column:

Visualize:

Lake Mead Levels  (1)

Time series of Lake Mead water levels:

Create a Tabular object:

Split the dates into year and month columns using TransformColumns and remove "Date" column:

Use PivotToColumns to convert into a more compact form for visualization:

Properties & Relations  (3)

TransformColumns keeps all the columns not being transformed:

ConstructColumns keeps only the listed columns:

TransformColumns can be used to remove columns:

Equivalently, use DeleteColumns:

Take a Tabular object whose columns are years and months:

TransformColumns can add a new column by constructing a date for each row:

Alternatively, first construct the column of dates and then use InsertColumns to append it:

Wolfram Research (2025), TransformColumns, Wolfram Language function, https://reference.wolfram.com/language/ref/TransformColumns.html.

Text

Wolfram Research (2025), TransformColumns, Wolfram Language function, https://reference.wolfram.com/language/ref/TransformColumns.html.

CMS

Wolfram Language. 2025. "TransformColumns." Wolfram Language & System Documentation Center. Wolfram Research. https://reference.wolfram.com/language/ref/TransformColumns.html.

APA

Wolfram Language. (2025). TransformColumns. Wolfram Language & System Documentation Center. Retrieved from https://reference.wolfram.com/language/ref/TransformColumns.html

BibTeX

@misc{reference.wolfram_2024_transformcolumns, author="Wolfram Research", title="{TransformColumns}", year="2025", howpublished="\url{https://reference.wolfram.com/language/ref/TransformColumns.html}", note=[Accessed: 15-January-2025 ]}

BibLaTeX

@online{reference.wolfram_2024_transformcolumns, organization={Wolfram Research}, title={TransformColumns}, year={2025}, url={https://reference.wolfram.com/language/ref/TransformColumns.html}, note=[Accessed: 15-January-2025 ]}