问在ValueError中放置“`Replace = True`”：当“replace=False”时不能获取比总体更大的样本

EN

# Split data
raw_train_df, valid_df = train_test_split(image_df, test_size = 0.25, random_state = 12345, stratify = 
image_df['class_name'])

# Print results
print(raw_train_df.shape, 'Training data')
print(valid_df.shape, 'Validation data')

(11250, 10) Training data
(3750, 10) Validation data

fig, (ax1, ax2) = plt.subplots(1, 2, figsize = (20, 10))
raw_train_df.groupby('class_name').size().plot.bar(ax = ax1)
train_df = raw_train_df.groupby('class_name').\
    apply(lambda x: x.sample(TRAIN_SAMPLES//15)).\ # Here I put 15 instead of 3, because I have 15 
classes
    reset_index(drop=True)
train_df.groupby('class_name').size().plot.bar(ax=ax2) 
print(train_df.shape[0], 'new training size')

ValueError                                Traceback (most recent call last)
<ipython-input-16-3b4d2b82246c> in <module>()
  1 fig, (ax1, ax2) = plt.subplots(1, 2, figsize = (20, 10))
  2 raw_train_df.groupby('class_name').size().plot.bar(ax = ax1)
----> 3 train_df = raw_train_df.groupby('class_name').    apply(lambda x: 
x.sample(TRAIN_SAMPLES//15)).    reset_index(drop=True)
  4 train_df.groupby('class_name').size().plot.bar(ax=ax2)
  5 print(train_df.shape[0], 'new training size')

4 frames
/usr/local/lib/python3.6/dist-packages/pandas/core/generic.py in sample(self, n, frac, replace, 
weights, random_state, axis)
4993             )
4994 
-> 4995         locs = rs.choice(axis_length, size=n, replace=replace, p=weights)
4996         return self.take(locs, axis=axis)
4997 

mtrand.pyx in numpy.random.mtrand.RandomState.choice()

ValueError: Cannot take a larger sample than population when 'replace=False'