Python 属性错误：'；非类型'；对象没有属性'；安装发电机和x27；

代码：

我得到的错误是：

回溯（最近一次呼叫最后一次）：

文件“”，第1行，在 型号。安装发电机（发电机组

AttributeError:“非类型”对象没有属性“适合生成器”

我已经导入了Keras.models，其中还包含fit_generator，并尝试使用fit代替fit_generator，但仍然无法解决此问题

---

期待您的帮助！

型号为无。请检查此行以了解原因：

import numpy as np 
import pandas as pd
import os
from tqdm import tqdm

# Fix seeds
from numpy.random import seed
seed(639)
from tensorflow import set_random_seed
set_random_seed(5944)

# Import
float_data = pd.read_csv("train.csv", dtype={"acoustic_data": np.float32, "time_to_failure": np.float32}).values

# Helper function for the data generator. Extracts mean, standard deviation, and quantiles per time step.
# Can easily be extended. Expects a two dimensional array.
def extract_features(z):
     return np.c_[z.mean(axis=1), 
                  z.min(axis=1),
                  z.max(axis=1),
                  z.std(axis=1)]

# For a given ending position "last_index", we split the last 150'000 values 
# of "x" into 150 pieces of length 1000 each. So n_steps * step_length should equal 150'000.
# From each piece, a set features are extracted. This results in a feature matrix 
# of dimension (150 time steps x features).  
def create_X(x, last_index=None, n_steps=150, step_length=1000):
    if last_index == None:
        last_index=len(x)

    assert last_index - n_steps * step_length >= 0

    # Reshaping and approximate standardization with mean 5 and std 3.
    temp = (x[(last_index - n_steps * step_length):last_index].reshape(n_steps, -1) - 5 ) / 3

    # Extracts features of sequences of full length 1000, of the last 100 values and finally also 
    # of the last 10 observations. 
    return np.c_[extract_features(temp),
                 extract_features(temp[:, -step_length // 10:]),
                 extract_features(temp[:, -step_length // 100:])]

# Query "create_X" to figure out the number of features
n_features = create_X(float_data[0:150000]).shape[1]
print("Our RNN is based on %i features"% n_features)

# The generator endlessly selects "batch_size" ending positions of sub-time series. For each ending position,
# the "time_to_failure" serves as target, while the features are created by the function "create_X".
def generator(data, min_index=0, max_index=None, batch_size=16, n_steps=150, step_length=1000):
    if max_index is None:
        max_index = len(data) - 1

    while True:
        # Pick indices of ending positions
        rows = np.random.randint(min_index + n_steps * step_length, max_index, size=batch_size)

        # Initialize feature matrices and targets
        samples = np.zeros((batch_size, n_steps, n_features))
        targets = np.zeros(batch_size, )

        for j, row in enumerate(rows):
            samples[j] = create_X(data[:, 0], last_index=row, n_steps=n_steps, step_length=step_length)
            targets[j] = data[row - 1, 1]
        yield samples, targets

batch_size = 64

# Position of second (of 16) earthquake. Used to have a clean split
# between train and validation
second_earthquake = 50085877
float_data[second_earthquake, 1]

# Initialize generators
train_gen = generator(float_data, batch_size=batch_size) # Use this for better score
# train_gen = generator(float_data, batch_size=batch_size, min_index=second_earthquake + 1)
valid_gen = generator(float_data, batch_size=batch_size, max_index=second_earthquake)

# Define model
from keras.models import Sequential
from keras.layers import Dense
from keras.optimizers import adam
from keras.callbacks import ModelCheckpoint
from keras.models import Model
from keras.callbacks import ModelCheckpoint
from keras.models import Sequential
from keras.layers import Dense, Activation, Flatten
from sklearn.model_selection import train_test_split
from sklearn.ensemble import RandomForestRegressor
from sklearn.metrics import mean_absolute_error 
from matplotlib import pyplot as plt
import matplotlib.pyplot as plt
import pandas as pd
import numpy as np
import warnings 



model = Sequential()
model.add(Dense(10, activation='relu'))
model.add(Dense(1))


cb = [ModelCheckpoint("model.hdf5", save_best_only=True, period=3)]
# Compile and fit model
model = model.compile(optimizer=adam(lr=0.0005), loss="mae")


history = model.fit_generator(train_gen,
                              steps_per_epoch=1000,
                              epochs=30,
                              verbose=0,
                              callbacks=cb,
                              validation_data=valid_gen,
                              validation_steps=200)
model.summary()
# Visualize accuracies
import matplotlib.pyplot as plt

def perf_plot(history, what = 'loss'):
    x = history.history[what]
    val_x = history.history['val_' + what]
    epochs = np.asarray(history.epoch) + 1

    plt.plot(epochs, x, 'bo', label = "Training " + what)
    plt.plot(epochs, val_x, 'b', label = "Validation " + what)
    plt.title("Training and validation " + what)
    plt.xlabel("Epochs")
    plt.legend()
    plt.show()
    return None

perf_plot(history)

# Load submission file
submission = pd.read_csv('sample_submission.csv', index_col='seg_id', dtype={"time_to_failure": np.float32})

# Load each test data, create the feature matrix, get numeric prediction
for i, seg_id in enumerate(tqdm(submission.index)):
  #  print(i)
    seg = pd.read_csv('../test/' + seg_id + '.csv')
    x = seg['acoustic_data'].values
    submission.time_to_failure[i] = model.predict(np.expand_dims(create_X(x), 0))

submission.head()

# Save
submission.to_csv('submissionearth.csv')

你的问题在这里：

model = model.compile(optimizer=adam(lr=0.0005), loss="mae")

您不应该分配

model.compile（…）

，因为它不会返回任何内容，相反，该行应该只读取

model.compile（optimizer=adam（lr=0.0005），loss=“mae”）

，所以只需将其设置为如下所示

model = model.compile(optimizer=adam(lr=0.0005), loss="mae")


history = model.fit_generator(train_gen,
                          steps_per_epoch=1000,
                          epochs=30,
                          verbose=0,
                          callbacks=cb,
                          validation_data=valid_gen,
                          validation_steps=200)

model.compile(optimizer=adam(lr=0.0005), loss="mae")


history = model.fit_generator(train_gen,
                          steps_per_epoch=1000,
                          epochs=30,
                          verbose=0,
                          callbacks=cb,
                          validation_data=valid_gen,
                          validation_steps=200)