Improved documentation of bucketized_column.
Change: 155389503
This commit is contained in:
parent
dcb6e92b1d
commit
ec8ffb9eaf
@ -393,6 +393,17 @@ def bucketized_column(source_column, boundaries):
|
|||||||
`boundaries=[0., 1., 2.]` generates buckets `(-inf, 0.)`, `[0., 1.)`,
|
`boundaries=[0., 1., 2.]` generates buckets `(-inf, 0.)`, `[0., 1.)`,
|
||||||
`[1., 2.)`, and `[2., +inf)`.
|
`[1., 2.)`, and `[2., +inf)`.
|
||||||
|
|
||||||
|
For example, if the inputs are
|
||||||
|
`boundaries` = [0, 10, 100]
|
||||||
|
input tensor = [[-5, 10000]
|
||||||
|
[150, 10]
|
||||||
|
[5, 100]]
|
||||||
|
|
||||||
|
then the output will be
|
||||||
|
output = [[0, 3]
|
||||||
|
[3, 2]
|
||||||
|
[1, 3]]
|
||||||
|
|
||||||
Example:
|
Example:
|
||||||
|
|
||||||
```python
|
```python
|
||||||
|
Loading…
Reference in New Issue
Block a user