Expands a list-column of numeric vectors into separate columns (when applied to
a tibble) or into a matrix (when applied to a list-column directly). This is the
inverse operation of fold().
Usage
unfold(x, ...)
# S3 method for class 'data.frame'
unfold(x, col, .prefix = NULL, ...)
# S3 method for class 'list'
unfold(x, .prefix = NULL, ...)Arguments
- x
A data frame, tibble, or list of numeric vectors
- ...
Additional arguments passed to methods
- col
Name of the list-column to unfold (unquoted). If missing, automatically detects the single coe column using
get_coe_cols().- .prefix
Character string to prefix column names. Default
NULLmeans no prefix. Set to a string to add a prefix to all column names.
Value
For tibbles: A tibble with the list-column removed and replaced by multiple numeric columns
For lists: A numeric matrix with one row per list element
Details
unfold() is an S3 generic with methods for:
Data frames/tibbles (
unfold.data.frame): User-facing, expands list-columns into separate columns with prefixes and class preservationLists (
unfold.list): Internal use, converts list of vectors directly to a matrix for statistical functions
For data frames (tibbles)
The function:
Auto-detects the coe column if not specified
Extracts the list-column and its classes
Stacks all vectors into a matrix (by rows)
Converts the matrix to a tibble with prefixed column names
Copies classes from the list-column to each new column
Removes the original list-column and binds the new columns
This is useful when you need to access individual coefficient values for statistical analysis, plotting, or further manipulation.
For lists
When applied directly to a list of numeric vectors, unfold() converts it to
a matrix by stacking vectors as rows. This is useful for internal operations
that need matrix input (e.g., prcomp(), lda()).
The function:
Finds the first non-NA element to determine structure
Extracts or generates column names
Stacks all vectors into a matrix (one row per vector)
Preserves names as column names
See also
unfold.data.frame(), unfold.list(), fold()
Examples
# For tibbles - see ?unfold.data.frame
df_folded <- tibble::tibble(
id = 1:3,
coe = list(c(A1=1, A2=4), c(A1=2, A2=5), c(A1=3, A2=6))
)
unfold(df_folded, coe)
#> # A tibble: 3 × 3
#> id coe_A1 coe_A2
#> <int> <dbl> <dbl>
#> 1 1 1 4
#> 2 2 2 5
#> 3 3 3 6
# For lists directly (returns matrix)
coe_list <- list(c(A1=1, A2=4), c(A1=2, A2=5), c(A1=3, A2=6))
unfold(coe_list)
#> A1 A2
#> [1,] 1 4
#> [2,] 2 5
#> [3,] 3 6
# A1 A2
# [1,] 1 4
# [2,] 2 5
# [3,] 3 6
# Start with folded data
df_folded <- tibble::tibble(
id = 1:3,
coe = list(
c(A1 = 1, A2 = 4, B1 = 7, B2 = 10),
c(A1 = 2, A2 = 5, B1 = 8, B2 = 11),
c(A1 = 3, A2 = 6, B1 = 9, B2 = 12)
)
)
class(df_folded$coe) <- c("eft", "coe", "list")
# Auto-detect coe column
df_unfolded <- unfold(df_folded)
# Or explicit
df_unfolded <- unfold(df_folded, coe)
# Unfold with default prefix (column name + "_")
df_unfolded
#> # A tibble: 3 × 5
#> id coe_A1 coe_A2 coe_B1 coe_B2
#> <int> <dbl> <dbl> <dbl> <dbl>
#> 1 1 1 4 7 10
#> 2 2 2 5 8 11
#> 3 3 3 6 9 12
# A tibble: 3 × 5
# id coe_A1 coe_A2 coe_B1 coe_B2
# <int> <dbl> <dbl> <dbl> <dbl>
# 1 1 1 4 7 10
# Each column inherits classes
class(df_unfolded$coe_A1)
#> [1] "numeric"
# [1] "eft" "coe"
# Unfold without prefix
unfold(df_folded, coe, .prefix = "")
#> # A tibble: 3 × 5
#> id A1 A2 B1 B2
#> <int> <dbl> <dbl> <dbl> <dbl>
#> 1 1 1 4 7 10
#> 2 2 2 5 8 11
#> 3 3 3 6 9 12
# Creates: A1, A2, B1, B2
# Custom prefix
unfold(df_folded, coe, .prefix = "harm_")
#> # A tibble: 3 × 5
#> id harm_A1 harm_A2 harm_B1 harm_B2
#> <int> <dbl> <dbl> <dbl> <dbl>
#> 1 1 1 4 7 10
#> 2 2 2 5 8 11
#> 3 3 3 6 9 12
# Direct list to matrix conversion
coe_list <- list(
c(A1 = 1, A2 = 4, B1 = 7),
c(A1 = 2, A2 = 5, B1 = 8),
c(A1 = 3, A2 = 6, B1 = 9)
)
mat <- unfold(coe_list)
mat
#> A1 A2 B1
#> [1,] 1 4 7
#> [2,] 2 5 8
#> [3,] 3 6 9
# A1 A2 B1
# [1,] 1 4 7
# [2,] 2 5 8
# [3,] 3 6 9
# With prefix
unfold(coe_list, .prefix = "coef_")
#> coef_A1 coef_A2 coef_B1
#> [1,] 1 4 7
#> [2,] 2 5 8
#> [3,] 3 6 9
# coef_A1 coef_A2 coef_B1
# [1,] 1 4 7