如何解决使用文件列表构建 tf.data 管道,即pickle 数据框
我无法为腌制数据帧列表构建 tf.data
管道(在 python 3.7.7 和 Windows 10 上使用 Tensorflow 2.1.0)。要开始使用以下代码:
import numpy as np
import pandas as pd
import os,pickle
import tensorflow as tf
from tensorflow.keras.preprocessing.sequence import pad_sequences
from sklearn.preprocessing import MultiLabelBinarizer
MAX_LEN = 8
NULL_VALUE = -1
# creating the sample dataframe
df = pd.DataFrame({'SOURCE': [[0,0],[0,1,2,2],1],[-1,-1,-1],1]],'TIME': [[0,7,7]],'DATE': [[0,'CAT_AF1': [[1,3,4],[1,4,3],'CAT_AF2': [[1,'PROD': [[1,'PROD_Y': [[1,[1],[7],[3],[0],[2]]})
FEATURES = [col for col in df.columns if col != 'PROD_Y']
# saving the same dataframe as multiple pickle files
os.mkdir('data')
os.mkdir(os.path.join('data','pickles'))
for n in range(10):
with open(os.path.join('data','pickles','pickle-n-{n}.pkl'.format(n=n)),'wb') as f:
pickle.dump(df,f,pickle.HIGHEST_PROTOCOL)
def read_pickles(filename):
with open(filename,'rb') as f:
df_ = pickle.load(f)
return df_
def pad_(x):
return pad_sequences([x],dtype=np.int32,maxlen=MAX_LEN,padding='pre',truncating='pre',value=NULL_VALUE)[0]
def parse_files(filename):
df_ = read_pickles(filename)
MLB = MultiLabelBinarizer(classes=list(set([j for i in df_['PROD_Y'].tolist() for j in i])))
MLB.fit(y=df_['PROD_Y'])
for col in FEATURES:
df_[col] = df_[col].apply(pad_)
df_['PROD_Y'] = MLB.transform(df_['PROD_Y']).tolist()
return dict(df_)
# return tf.data.Dataset.from_tensor_slices(dict(df_))
# return df_.to_dict('list')
def parse_(filename):
out = tf.py_function(parse_files,inp=[filename],Tout=[tf.float32 for k in FEATURES]) # not sure how to write the dict returned into Tout argument
return out
list_of_files = [os.path.join('data',f) for f in os.listdir(os.path.join('data','pickles'))]
dataset = tf.data.Dataset.from_tensor_slices(list_of_files)
dataset_2 = dataset.map(parse_)
a = [d for d in dataset_2]
在最后一行,它在 read_pickles
中抛出以下错误:
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xb0 in position 0: invalid start byte
我尝试在 filename.decode('utf-8')
中使用 read_pickles
但这也会导致另一个错误:
AttributeError: 'tensorflow.python.framework.ops.EagerTensor' object has no attribute 'decode'
当我搜索它时,我找到了这些,但它们对问题没有帮助:
完整回溯:
2021-05-16 19:59:01.568800: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudart64_101.dll
2021-05-16 19:59:04.926696: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library nvcuda.dll
2021-05-16 19:59:06.100711: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1555] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: GeForce GTX 1660 Ti computeCapability: 7.5
coreClock: 1.59GHz coreCount: 24 deviceMemorySize: 6.00GiB deviceMemoryBandwidth: 268.26GiB/s
2021-05-16 19:59:06.100787: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudart64_101.dll
2021-05-16 19:59:06.109139: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cublas64_10.dll
2021-05-16 19:59:06.114618: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cufft64_10.dll
2021-05-16 19:59:06.117511: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library curand64_10.dll
2021-05-16 19:59:06.124338: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cusolver64_10.dll
2021-05-16 19:59:06.129912: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cusparse64_10.dll
2021-05-16 19:59:06.142492: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudnn64_7.dll
2021-05-16 19:59:06.142600: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1697] Adding visible gpu devices: 0
2021-05-16 19:59:06.143031: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2
2021-05-16 19:59:06.144206: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1555] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: GeForce GTX 1660 Ti computeCapability: 7.5
coreClock: 1.59GHz coreCount: 24 deviceMemorySize: 6.00GiB deviceMemoryBandwidth: 268.26GiB/s
2021-05-16 19:59:06.144245: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudart64_101.dll
2021-05-16 19:59:06.144265: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cublas64_10.dll
2021-05-16 19:59:06.144281: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cufft64_10.dll
2021-05-16 19:59:06.144299: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library curand64_10.dll
2021-05-16 19:59:06.144314: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cusolver64_10.dll
2021-05-16 19:59:06.144328: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cusparse64_10.dll
2021-05-16 19:59:06.144343: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudnn64_7.dll
2021-05-16 19:59:06.144390: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1697] Adding visible gpu devices: 0
2021-05-16 19:59:06.866245: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1096] Device interconnect StreamExecutor with strength 1 edge matrix:
2021-05-16 19:59:06.866285: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1102] 0
2021-05-16 19:59:06.866300: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] 0: N
2021-05-16 19:59:06.866531: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1241] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 4755 MB memory) -> physical GPU (device: 0,name: GeForce GTX 1660 Ti,pci bus id: 0000:01:00.0,compute capability: 7.5)
2021-05-16 19:59:07.063641: W tensorflow/core/framework/op_kernel.cc:1643] Invalid argument: UnicodeDecodeError: 'utf-8' codec can't decode byte 0xb0 in position 0: invalid start byte
Traceback (most recent call last):
File "C:\Users\D\Miniconda3\envs\Dev\lib\site-packages\tensorflow_core\python\ops\script_ops.py",line 234,in __call__
return func(device,token,args)
File "C:\Users\D\Miniconda3\envs\Dev\lib\site-packages\tensorflow_core\python\ops\script_ops.py",line 123,in __call__
ret = self._func(*args)
File "stack-overflow-tf-data-question.py",line 33,in parse_files
df_ = read_pickles(filename)
File "stack-overflow-tf-data-question.py",line 25,in read_pickles
with open(filename,'rb') as f:
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xb0 in position 0: invalid start byte
2021-05-16 19:59:07.063812: W tensorflow/core/framework/op_kernel.cc:1655] OP_REQUIRES failed at iterator_ops.cc:941 : Invalid argument: UnicodeDecodeError: 'utf-8' codec can't decode byte 0xb0 in position 0: invalid start byte
Traceback (most recent call last):
File "C:\Users\D\Miniconda3\envs\Dev\lib\site-packages\tensorflow_core\python\ops\script_ops.py",'rb') as f:
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xb0 in position 0: invalid start byte
[[{{node EagerPyFunc}}]]
stack-overflow-tf-data-question.py:25: DeprecationWarning: path should be string,bytes,or os.PathLike,not tensorflow.python.framework.ops.EagerTensor
with open(filename,'rb') as f:
Traceback (most recent call last):
File "C:\Users\D\Miniconda3\envs\Dev\lib\site-packages\tensorflow_core\python\eager\context.py",line 1897,in execution_mode
yield
File "C:\Users\D\Miniconda3\envs\Dev\lib\site-packages\tensorflow_core\python\data\ops\iterator_ops.py",line 659,in _next_internal
output_shapes=self._flat_output_shapes)
File "C:\Users\D\Miniconda3\envs\Dev\lib\site-packages\tensorflow_core\python\ops\gen_dataset_ops.py",line 2479,in iterator_get_next_sync
_ops.raise_from_not_ok_status(e,name)
File "C:\Users\D\Miniconda3\envs\Dev\lib\site-packages\tensorflow_core\python\framework\ops.py",line 6606,in raise_from_not_ok_status
six.raise_from(core._status_to_exception(e.code,message),None)
File "<string>",line 3,in raise_from
tensorflow.python.framework.errors_impl.InvalidArgumentError: UnicodeDecodeError: 'utf-8' codec can't decode byte 0xb0 in position 0: invalid start byte
Traceback (most recent call last):
File "C:\Users\D\Miniconda3\envs\Dev\lib\site-packages\tensorflow_core\python\ops\script_ops.py",'rb') as f:
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xb0 in position 0: invalid start byte
[[{{node EagerPyFunc}}]] [Op:IteratorGetNextSync]
During handling of the above exception,another exception occurred:
Traceback (most recent call last):
File "stack-overflow-tf-data-question.py",line 52,in <module>
a = [d for d in dataset_2]
File "stack-overflow-tf-data-question.py",in <listcomp>
a = [d for d in dataset_2]
File "C:\Users\D\Miniconda3\envs\Dev\lib\site-packages\tensorflow_core\python\data\ops\iterator_ops.py",line 630,in __next__
return self.next()
File "C:\Users\D\Miniconda3\envs\Dev\lib\site-packages\tensorflow_core\python\data\ops\iterator_ops.py",line 674,in next
return self._next_internal()
File "C:\Users\D\Miniconda3\envs\Dev\lib\site-packages\tensorflow_core\python\data\ops\iterator_ops.py",line 665,in _next_internal
return structure.from_compatible_tensor_list(self._element_spec,ret)
File "C:\Users\D\Miniconda3\envs\Dev\lib\contextlib.py",line 130,in __exit__
self.gen.throw(type,value,traceback)
File "C:\Users\D\Miniconda3\envs\Dev\lib\site-packages\tensorflow_core\python\eager\context.py",line 1900,in execution_mode
executor_new.wait()
File "C:\Users\D\Miniconda3\envs\Dev\lib\site-packages\tensorflow_core\python\eager\executor.py",line 67,in wait
pywrap_tensorflow.TFE_ExecutorWaitForAllPendingNodes(self._handle)
tensorflow.python.framework.errors_impl.InvalidArgumentError: UnicodeDecodeError: 'utf-8' codec can't decode byte 0xb0 in position 0: invalid start byte
Traceback (most recent call last):
File "C:\Users\D\Miniconda3\envs\Dev\lib\site-packages\tensorflow_core\python\ops\script_ops.py",'rb') as f:
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xb0 in position 0: invalid start byte
[[{{node EagerPyFunc}}]]
感谢您的帮助! TIA!
解决方法
将 read_pickles 函数替换为
def read_pickles(filename):
with open(filename,'rb',encoding="utf8") as f:
df_ = pickle.load(f)
return df_
版权声明:本文内容由互联网用户自发贡献,该文观点与技术仅代表作者本人。本站仅提供信息存储空间服务,不拥有所有权,不承担相关法律责任。如发现本站有涉嫌侵权/违法违规的内容, 请发送邮件至 dio@foxmail.com 举报,一经查实,本站将立刻删除。