Java 类名：com.alibaba.alink.operator.stream.image.WriteTensorToImageStreamOp
Python 类名：WriteTensorToImageStreamOp

功能介绍

将张量列转换为图片，并写入根目录对应的相对路径列中，然后原样输出结果。

参数说明

| 名称 | 中文名称 | 描述 | 类型 | 是否必须？ | 取值范围 | 默认值 | | —- | —- | —- | —- | —- | —- | —- |

| relativeFilePathCol | 文件路径列 | 文件路径列 | String | ✓ | 所选列类型为 [STRING] | |

| rootFilePath | 文件路径 | 文件路径 | String | ✓ | | |

| tensorCol | tensor列 | tensor列 | String | ✓ | 所选列类型为 [BOOL_TENSOR, BYTE_TENSOR, DOUBLE_TENSOR, FLOAT_TENSOR, INT_TENSOR, LONG_TENSOR, STRING, STRING_TENSOR, TENSOR, UBYTE_TENSOR] | |

| imageType | 图片类型 | 图片类型 | String | | “PNG”, “JPEG” | “PNG” |

| reservedCols | 算法保留列名 | 算法保留列 | String[] | | | null |

代码示例

Python 代码

  1. df_data = pd.DataFrame([
  2.     'sphx_glr_plot_scripted_tensor_transforms_001.png'
  3. ])
  5. stream_data = StreamOperator.fromDataframe(df_data, schemaStr = 'path string')
  7. readImageToTensorStreamOp = ReadImageToTensorStreamOp()\
  8.     .setRootFilePath("https://pytorch.org/vision/stable/_images/")\
  9.     .setRelativeFilePathCol("path")\
  10.     .setOutputCol("tensor")
  12. writeTensorToImageStreamOp = WriteTensorToImageStreamOp()\
  13.             .setRootFilePath("/tmp/write_tensor_to_image")\
  14.             .setTensorCol("tensor")\
  15.             .setImageType("png")\
  16.             .setRelativeFilePathCol("path")
  18. stream_data.link(readImageToTensorStreamOp).link(writeTensorToImageStreamOp).print()
  20. StreamOperator.execute()

Java 代码

  1. import org.apache.flink.types.Row;
  3. import com.alibaba.alink.operator.stream.StreamOperator;
  4. import com.alibaba.alink.operator.stream.source.MemSourceStreamOp;
  5. import com.alibaba.alink.params.image.HasImageType.ImageType;
  6. import org.junit.Test;
  8. import java.util.Collections;
  9. import java.util.List;
  11. public class WriteTensorToImageStreamOpTest {
  13.     @Test
  14.     public void testWriteTensorToImageStreamOp() throws Exception {
  16.         List <Row> data = Collections.singletonList(
  17.             Row.of("sphx_glr_plot_scripted_tensor_transforms_001.png")
  18.         );
  20.         MemSourceStreamOp memSourceStreamOp = new MemSourceStreamOp(data, "path string");
  22.         ReadImageToTensorStreamOp readImageToTensorStreamOp = new ReadImageToTensorStreamOp()
  23.             .setRootFilePath("https://pytorch.org/vision/stable/_images/")
  24.             .setRelativeFilePathCol("path")
  25.             .setOutputCol("tensor");
  27.         WriteTensorToImageStreamOp writeTensorToImageStreamOp = new WriteTensorToImageStreamOp()
  28.             .setRootFilePath("/tmp/write_tensor_to_image")
  29.             .setTensorCol("tensor")
  30.             .setImageType(ImageType.PNG)
  31.             .setRelativeFilePathCol("path");
  33.         memSourceStreamOp.link(readImageToTensorStreamOp).link(writeTensorToImageStreamOp).print();
  35.         StreamOperator.execute();
  36.     }
  37. }

运行结果

可以在 /tmp/write_tensor_to_image/sphx_glr_plot_scripted_tensor_transforms_001.png 中找到 https://pytorch.org/vision/stable/_images/sphx_glr_plot_scripted_tensor_transforms_001.png
同时组件的输出结果为：
| path | tensor |
|—————————————————————————+————————————————|
| sphx_glr_plot_scripted_tensor_transforms_001.png | FLOAT#250,520,4#255.0 255.0… |