HistogramFixedWidth
English | 简体中文
Product Support Status
| Product | Supported |
|---|---|
| Ascend 950PR&950DT products | √ |
| Atlas A3 products | √ |
| Atlas A2 products | √ |
| Atlas 200I/500 A2 inference products | × |
| Atlas inference products | √ |
| Atlas training products | × |
Function Description
- Operator function: computes the histogram of a tensor. This operator is compatible with the histogram_fixed_width operator of TensorFlow.
- Computation formula: uses range=[min,max] as the upper and lower bounds of the statistics, divides the space between min and max into nbins intervals of equal width, and counts the number of elements in tensor x in each interval. Elements smaller than min are counted into the first interval, and elements greater than max are counted into the last interval.
Parameter Description
| Parameter Name | Input/Output/Attribute | Description | Data Type | Data Format |
|---|---|---|---|---|
| x | Input | Input tensor, which is the data to be counted. | FLOAT, FLOAT16, INT32, INT64 | ND |
| range | Input | Tensor with the shape [2], containing [min,max]. | FLOAT, FLOAT16, INT32, INT64 | ND |
| nbins | Input | Scalar tensor, which is the number of histogram intervals. | INT32 | ND |
| y | Output | Histogram result, with the shape [nbins]. | INT32 | ND |
| dtype | Attribute | Reserved attribute, and the default value is 3 | INT64 | - |
Constraint Description
The max of the input tensor range must be greater than min. Otherwise, an error is reported.
Invocation Description
| Invocation Method | Invocation Sample | Description |
|---|---|---|
| Graph-mode invocation | test_geir_histogram_fixed_width | Invokes the HistogramFixedWidth operator by building a graph through the operator IR. |