Dataset pack
Training and test values with gaps
A numeric feature split into training and test sets, with missing values in both.
Fictional data · UTF-8 CSV, every value quoted so padding and leading zeros survive; empty fields are missing values · free for personal and classroom practice · practise it in Fill gaps with a median learned from training data only.
train 6 rows × 2 columns
Download train.csv (Pro)Data dictionary
- id
- Row ID.
- x
- Whole number or missing.
test 8 rows × 2 columns
Download test.csv (Pro)Data dictionary
- id
- Row ID.
- x
- Whole number or missing; extreme values occur.
Values are shown in quotes so leading and trailing spaces are visible. ∅ means the cell is missing. Exercises are also graded on hidden datasets that are never published.