我从protobuf文件中读取消息。 该消息包含时间序列数据,因此,我预期类似矩阵的结构,例如processed_row [nData,nVar]
print(mySerializedData.processed_row)
>> [timestamp: 0.0
linear_acc_x: 0.288501300049
linear_acc_y: 0.573411297607
linear_acc_z: 0.161608612061
, timestamp: 0.0049
linear_acc_x: 0.428562097168
linear_acc_y: 0.685938775635
linear_acc_z: 0.221463653564
, timestamp: 0.01
linear_acc_x: 0.45968671875
linear_acc_y: 0.738611212158
linear_acc_z: 0.185550628662]
我可以访问个人数据,如
print(mySerializedData.processed_row[0].timestamp)
>> 0.0
print(mySerializedData.processed_row[1].timestamp)
>> 0.0049
但我想要获得的是
print(mySerializedData.processed_row[:].timestamp)
但它显示错误AttributeError: 'list' object has no attribute 'timestamp'
print(type(mySerializedData.processed_row[0].timestamp))
>> <type 'float'>
print(type(mySerializedData.processed_row[0]))
>> <class 'PushCore_pb2.ProcessedDataRow'>
有没有办法像[[]]一样得到timestamp
双数组?
谢谢
答案 0 :(得分:0)
您可以使用list comprehension:
创建时间戳值列表[element.timestamp for element in mySerializedData.processed_row]
这是一个概括性的例子。在example.proto
:
package example;
message Element {
required double timestamp = 1;
required string data = 2;
}
message Container {
repeated Element elements = 1;
}
使用:
生成相应的Python输出protoc --python_out=. example.proto
然后,在Python中:
>>> from example_pb2 import Container, Element
>>>
>>> container = Container(
... elements=[
... Element(timestamp=.1, data='Some data.'),
... Element(timestamp=.2, data='Some more data.'),
... ]
... )
>>>
>>> [element.timestamp for element in container.elements]
[0.1, 0.2]