Functional iterator for single axis for Tensor (not TensorView) #3
Conversation
![github-advanced-security[bot]](https://avatars.githubusercontent.com/in/57789?s=60&v=4){kind=small}
| @@ -0,0 +1,116 @@ | |||
| use crate::data_buffer::{DataBuffer, DataOwned, DataView}; | |||
Check warning
Code scanning / clippy
unused import: DataOwned Warning
| @@ -0,0 +1,116 @@ | |||
| use crate::data_buffer::{DataBuffer, DataOwned, DataView}; | |||
| use crate::dtype::RawDataType; | |||
| use crate::{tensor, TensorView}; | |||
Check warning
Code scanning / clippy
unused import: tensor Warning
| } | ||
| } | ||
|
|
||
| impl<'a, T, B> Iterator for IteratorBase<'a, T, B> |
Check warning
Code scanning / clippy
the following explicit lifetimes could be elided: 'a Warning
| } | ||
| } | ||
|
|
||
| impl<'a, T, B> Iterator for IteratorBase<'a, T, B> |
Check warning
Code scanning / clippy
the following explicit lifetimes could be elided: 'a Warning
| let mut new_shape = self.shape.clone(); | ||
| let mut new_stride = self.stride.clone(); | ||
|
|
||
| for i in 0..self.axis { |
Check warning
Code scanning / clippy
the loop variable i is only used to index new_stride Warning
| let mut new_stride = self.stride.clone(); | ||
|
|
||
| for i in 0..self.axis { | ||
| new_stride[i] = new_stride[i] / new_shape[self.axis]; |
Check warning
Code scanning / clippy
manual implementation of an assign operation Warning
This iterator works for a single axis and on a Tensor object. Users can access the shape, stride, and elements of each iterable as a normal TensorView object. Testing has been done for 2, 3, and 4-dimensional arrays across multiple axes. In each new test created, only one axis has been uncommented, while others are left commented for ease of comprehension. For now, testing has been done manually.