Package-level declarations

Accessing and specifying columns is one of the most important parts of the API, used in most DataFrame operations.

DataFrame cannot contain columns with duplicate names. However, sometimes after reading dataframes from sources or after some operations, columns with duplicate names may appear in the result.

In many DSLs, the lambda expr {} can be used to create a new column by defining an expression to fill up each row.

If you've come across notations like a((b| [c, .. ] ) ) either in the KDocs or on the website and would like some further explanation for what it means, you've come to the right place.

Any Documentable annotated with this annotation will be excluded from the generated sources by the documentation processor.

Any Documentable annotated with this annotation will be exported to HTML by the documentation processor.

Expressing values using a "DataFrame Expression" can occur exclusively in a DataFrame Expression.

Expressing values using a "Row Expression" (See Row Expressions on the documentation website.) can occur in the following two types of operations:

Expressing values using a "Row-Column Expression" can occur exclusively in a Row Column Expression.

NA in DataFrame can be seen as "NaN or null".

Floats or Doubles can be represented as Float.NaN or Double.NaN, respectively, in cases where a mathematical operation is undefined, such as dividing by zero.

Selecting columns for various DataFrame operations can be done in the following ways:

Selecting rows that satisfy a "Row Condition" (See Row Conditions on the documentation website.) can occur in the following two types of operations:

The concept of unifying numbers is converting them to a common number type without losing information.