shape: (5, 2)
| Student | Score |
|---|---|
| str | i64 |
| "Harvey" | 88 |
| "Donna" | 76 |
| "Mike" | 93 |
| "Jessica" | 85 |
| "Louis" | 91 |
Get top values without sorting in polars
top_k
100DaysOfPolars
Sorting values is an expensive operation, especially when you have a large dataset. However, sometimes you have no choice but to sort the data, particularly when you need to pick the top 5 or 10 values. Below is a dataframe showing student scores on a test.
Get top 3 scores
Suppose you want to get the three students with the highest scores without sorting the data first. How can you do it? Polars has the method top_k, which you can use to accomplish exactly that.
(df
.top_k(3, by='Score')
)
shape: (3, 2)
| Student | Score |
|---|---|
| str | i64 |
| "Mike" | 93 |
| "Louis" | 91 |
| "Harvey" | 88 |
Tip
If you want to get the bottom 3 students, use bottom_k.
Sign up for my Polars course to improve your data analysis skills.