请问模型能对单变量时间序列数据进行预测吗？

[Jeff-Zhu-520]

我想对比一下单变量和多变量之间的差距，如何在参数中设置对单变量时间序列数据进行预测呢？

[YoZhibo]

您好，可以试试把features参数修改为S（单变量预测），enc_in，dec_in，c_out等参数修改为1。可以参考data_loader中的数据集类去适配您的数据集。

[Jeff-Zhu-520]

请问做单变量预测，总是出现矩阵维度不匹配的问题是怎么回事？
d_ff=128, moving_avg=25, factor=3, distil=True, dropout=0.05, embed='timeF', activation='gelu', output_attention=False, do_predict=False, num_workers=8, itr=1, train_epochs=10, batch_size=32, patience=3, learning_rate=0.0001, des='dou2 Prediction Exp', loss='MSE', lradj='type1', use_amp=False, use_gpu=True, gpu=0, use_multi_gpu=False, devices='0,1,2,3', test_flop=False)
Use GPU: cuda:0

start training : dou2_24_1_MSGNet_custom_ftS_sl24_ll12_pl1_dm64_nh8_el2_dl1_df128_fc3_ebtimeF_dtTrue_dou2 Prediction Exp_0>>>>>>>>>>>>>>>>>>>>>>>>>>
train 825
val 123
test 242
Traceback (most recent call last):
File "D:\MSGNet-main\MSGNet-main\run_longExp.py", line 154, in
exp.train(setting)
File "D:\MSGNet-main\MSGNet-main\exp\exp_main.py", line 156, in train
outputs = self.model(batch_x, batch_x_mark, dec_inp, batch_y_mark)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\module.py", line 1736, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\module.py", line 1747, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "D:\MSGNet-main\MSGNet-main\models\MSGNet.py", line 122, in forward
enc_out = self.layer_norm(self.modeli)
^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\module.py", line 1736, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\module.py", line 1747, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "D:\MSGNet-main\MSGNet-main\models\MSGNet.py", line 47, in forward
x = self.gconvi
^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\module.py", line 1736, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\module.py", line 1747, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "D:\MSGNet-main\MSGNet-main\layers\MSGBlock.py", line 151, in forward
out = self.linear(out)
^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\module.py", line 1736, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\module.py", line 1747, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\linear.py", line 125, in forward
return F.linear(input, self.weight, self.bias)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: mat1 and mat2 shapes cannot be multiplied (32x24 and 1x64)

[stu-github]

请问这个问题解决了吗？我也在做单变量预测

请问做单变量预测，总是出现矩阵维度不匹配的问题是怎么回事？ d_ff=128, moving_avg=25, factor=3, distil=True, dropout=0.05, embed='timeF', activation='gelu', output_attention=False, do_predict=False, num_workers=8, itr=1, train_epochs=10, batch_size=32, patience=3, learning_rate=0.0001, des='dou2 Prediction Exp', loss='MSE', lradj='type1', use_amp=False, use_gpu=True, gpu=0, use_multi_gpu=False, devices='0,1,2,3', test_flop=False) Use GPU: cuda:0

start training : dou2_24_1_MSGNet_custom_ftS_sl24_ll12_pl1_dm64_nh8_el2_dl1_df128_fc3_ebtimeF_dtTrue_dou2 Prediction Exp_0>>>>>>>>>>>>>>>>>>>>>>>>>>
train 825
val 123
test 242
Traceback (most recent call last):
File "D:\MSGNet-main\MSGNet-main\run_longExp.py", line 154, in
exp.train(setting)
File "D:\MSGNet-main\MSGNet-main\exp\exp_main.py", line 156, in train
outputs = self.model(batch_x, batch_x_mark, dec_inp, batch_y_mark)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\module.py", line 1736, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\module.py", line 1747, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "D:\MSGNet-main\MSGNet-main\models\MSGNet.py", line 122, in forward
enc_out = self.layer_norm(self.modeli)
^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\module.py", line 1736, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\module.py", line 1747, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "D:\MSGNet-main\MSGNet-main\models\MSGNet.py", line 47, in forward
x = self.gconvi
^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\module.py", line 1736, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\module.py", line 1747, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "D:\MSGNet-main\MSGNet-main\layers\MSGBlock.py", line 151, in forward
out = self.linear(out)
^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\module.py", line 1736, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\module.py", line 1747, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Jeff.conda\envs\test\Lib\site-packages\torch\nn\modules\linear.py", line 125, in forward
return F.linear(input, self.weight, self.bias)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: mat1 and mat2 shapes cannot be multiplied (32x24 and 1x64)