ValueCounts / Computation Layer

Count occurrences of unique values in a column, returning two new columns: one for unique values and another for their frequencies. Similar to pandas value_counts(), Polars value_counts(), or R's table() function.

Mathematical form: For each unique value $v_{i}$ in column $X$ : $f_{i} = j = 1 \sum n [X_{j} = v_{i}]$ When normalized: $p_{i} = \frac{f _{i}}{\sum _{k} f _{k}}$

Example with raw counts:

value	count
blue	3
red	2
green	1

Example with normalization:

value	proportion
blue	0.50
red	0.33
green	0.17

Common applications:

Frequency analysis
Category distribution
Data profiling
Outlier detection
Feature engineering
Quality checks
Distribution analysis

Note: Null values are excluded from counting by default. The output will be added as two new columns to the dataframe.

Table

Select

column

The column to analyze. Common input types:

Categorical variables (product types, status codes)
Discrete numeric values (ratings, counts)
Text data (tags, labels)
Binary flags (true/false values)

Maximum column name length: 128 characters

Sort

bool

false

Control output ordering. Examples:

With sort=true:

value	count
blue	3
red	2
green	1

With sort=false (arbitrary order):

value	count
red	2
green	1
blue	3

Normalize

bool

false

Control output values. Examples:

With normalize=false (counts):

value	count
blue	3
red	2
green	1

With normalize=true (proportions):

value	proportion
blue	0.50
red	0.33
green	0.17