FI_test: Imputation Test Data

In FastImputation: Learn from Training Data then Quickly Fill in Missing Data

Imputation Test Data

Description

Smaller simulated dataset drawn from the same distribution as FI_train and FI_true. This dataset is entirely the same as FI_true except this one has 5% of its values missing. Used with FastImputation.

Usage

data(FI_test)

Format

A data frame with 9 variables and 250 observations.

user_id_1

Sequential user ids

bounded_below_2

Multivariate normal, transformed using exp(x)

unbounded_3

Multivariate normal

unbounded_4

Multivariate normal

bounded_above_5

Multivariate normal, transformed using -exp(x)

bounded_above_and_below_6

Multivariate normal, transformed using pnorm(x)

unbounded_7

Multivariate normal

unbounded_8

Multivariate normal

categorical_9

"A" if the first of three multivariate normal draws is greatest; "B" if the second is greatest; "C" if the third is greatest

Author(s)

Stephen R. Haptonstahl [e-mail]

Source

All columns start as multivariate normal draws. Columns 2, 5, and 6 are transformed. Column 9 is the result of three multivariate normal columns being interpreted as one-hot encoding of a three-valued categorical variable.

FastImputation documentation built on Sept. 25, 2023, 5:06 p.m.