Das deutsche Python-Forum

Hey, bin absoluter Neuling und bekomme beim ausführen immer die beiden oben genannten Fehlermeldungen quasi im Wechsel angezeigt. Mal bekomme ich als Fehler "Can't handle a mix of multilabel-indicator and binary targets" angezeigt und mal "can't handle a mix of multilabel-indicator and multiclass " ohne, dass ich am Code irgendwas verändere..
Beim googlen stoße ich immer wieder auf np.argmax als Heilbringer, aber in meinem Fall scheint es keine Hilfe zu sein.
Habe jetzt nicht den gesamten Code eingefügt, nur das was evtl wichtig sein könnte.

Code: Alles auswählen

class DataManager(object):
    def __init__(self):
        self.train = DataLoader("xyz").load()  # path for train dataset
        self.test = DataLoader("xyzt").load()  # path for test dataset
data = DataManager()
input_train = data.train.sampledata
input_train = input_train.reshape(-1,  375, 1)
target_train = data.train.labeldata
target_train = target_train.reshape(-1, 1).astype('int32')
target_train = one_hot(target_train[:, -1], no_classes)
input_test = data.test.sampledata
input_test = input_test.reshape(-1,  375, 1)
target_test = data.test.labeldata
target_test = target_test.reshape(-1, 1).astype('int32')
target_test = one_hot(target_test[:, -1], no_classes)

pred_test = model.predict(input_test)
pred_test = np.argmax(pred_test, axis =1)

f1score = sklearn.metrics.f1_score(target_test, pred_test)
print('f1score:' + {f1score})

Bin dankbar für jede Hilfe

pred_test
[ 0 0 0 0 0 0 19 19 0 0]
pred_test-Shape
(1000,)
target_test
[[1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]
[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]
[1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]
[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]
[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]
[1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]
[1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]
[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]
[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]
[1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]]
target_test_shape
(1000, 21)
input_test
[[[0.342857]
[0.342857]
[0.35 ]
...
[0. ]
[0. ]
[0. ]]]
input_test_shape
(1000, 375, 1)

Hi, bin absoluter Neuling und bekomme beim ausführen meines Codes immer die oben genannte Fehlermeldung. Beim googlen stoße ich immer wieder auf den Heilbringer np.argmax, aber dies löst mein Problem anscheinend nicht.
Anbei mal paar Auszüge des Codes + Fehlermeldung.

Code: Alles auswählen

 class DataManager(object):
    def __init__(self):
        self.train = DataLoader("xyz/alltrain").load()  # path for train dataset
        self.test = DataLoader("xyz/alltest").load()  # path for test dataset
data = DataManager()
input_train = data.train.sampledata
input_train = input_train.reshape(-1,  375, 1)
target_train = data.train.labeldata
target_train = target_train.reshape(-1, 1).astype('int32')
target_train = one_hot(target_train[:, -1], no_classes)
input_test = data.test.sampledata
input_test = input_test.reshape(-1,  375, 1)
target_test = data.test.labeldata
target_test = target_test.reshape(-1, 1).astype('int32')
target_test = one_hot(target_test[:, -1], no_classes)
print(input_train.shape)
model = models.Sequential()

pred_test = model.predict(input_test)
pred_test = np.argmax(pred_test, axis =1)

print("pred_test")
print(pred_test[:10])
print("pred_test-Shape")
print(pred_test.shape)
print("target_test")
print(target_test[:10])
print("target_test_shape")
print(target_test.shape)



f1score = sklearn.metrics.f1_score(target_test, pred_test)
print('f1score:' + {f1score})

Fehlermeldungen :
Traceback (most recent call last):
File "C:/Users/derg/PycharmProjects/ecgcnnmulticlass/main.py", line 160, in <module>
f1score = sklearn.metrics.f1_score(target_test, pred_test)
File "C:\Users\derg\anaconda3\envs\Tensorflow2\lib\site-packages\sklearn\utils\validation.py", line 72, in inner_f
return f(**kwargs)
File "C:\Users\derg\anaconda3\envs\Tensorflow2\lib\site-packages\sklearn\metrics\_classification.py", line 1044, in f1_score
return fbeta_score(y_true, y_pred, beta=1, labels=labels,
File "C:\Users\derg\anaconda3\envs\Tensorflow2\lib\site-packages\sklearn\utils\validation.py", line 72, in inner_f
return f(**kwargs)
File "C:\Users\derg\anaconda3\envs\Tensorflow2\lib\site-packages\sklearn\metrics\_classification.py", line 1168, in fbeta_score
_, _, f, _ = precision_recall_fscore_support(y_true, y_pred,
File "C:\Users\derg\anaconda3\envs\Tensorflow2\lib\site-packages\sklearn\utils\validation.py", line 72, in inner_f
return f(**kwargs)
File "C:\Users\derg\anaconda3\envs\Tensorflow2\lib\site-packages\sklearn\metrics\_classification.py", line 1433, in precision_recall_fscore_support
labels = _check_set_wise_labels(y_true, y_pred, average, labels,
File "C:\Users\derg\anaconda3\envs\Tensorflow2\lib\site-packages\sklearn\metrics\_classification.py", line 1250, in _check_set_wise_labels
y_type, y_true, y_pred = _check_targets(y_true, y_pred)
File "C:\Users\derg\anaconda3\envs\Tensorflow2\lib\site-packages\sklearn\metrics\_classification.py", line 90, in _check_targets
raise ValueError("Classification metrics can't handle a mix of {0} "
ValueError: Classification metrics can't handle a mix of multilabel-indicator and multiclass targets

@Koerninger: bitte nicht zum gleichen Thema wieder einen neuen Post erstellen. Du hast die gleiche Frage schon hier viewtopic.php?f=1&t=53487&p=396783#p396783 gestellt.

Das deutsche Python-Forum

Can't handle a mix of multilabel-indicator and binary targets/can't handle a mix of multilabel-indicator and multiclass

Can't handle a mix of multilabel-indicator and binary targets/can't handle a mix of multilabel-indicator and multiclass

Re: Can't handle a mix of multilabel-indicator and binary targets/can't handle a mix of multilabel-indicator and multicl

ValueError: Classification metrics can't handle a mix of multilabel-indicator and multiclass targets

Re: ValueError: Classification metrics can't handle a mix of multilabel-indicator and multiclass targets