这是我的代码:

used_time_period = "2009-01-01::2017-04-01"data_used = data_input[used_time_period,]split_coefficient = 0.8train_set_rate = round(nrow(data_used) * split_coefficient)data_train = data_used[1:train_set_rate,]data_test = data_used[(train_set_rate + 1):nrow(data_used),]model = keras_model_sequential() %>%layer_simple_rnn(units = 75, input_shape = dim(data_train[,1:3]), activation = "relu", return_sequences = TRUE) %>% layer_dense(units = 2, activation = "relu")model %>% compile(optimizer = "adam", loss = "binary_crossentropy", metrics = "binary_accuracy")history = model %>% fit(x = data_train[,1:3], y = data_train[,4:5], epochs = 40, batch_size = 20)

我得到的错误是这样的:

ValueError: 检查输入时出错：期望 simple_rnn_input 具有3个维度，但得到的数组形状为(1661, 3)

dim(data_train[,1:3]) = (1661, 3)

dim(data_train[,4:5]) = (1661, 2)

我哪里做错了?

回答：

正如错误消息所说，layer_simple_rnn需要一个3D数组，但你使用的是一个data.frame，这是一个2D数组（一个带有行和列的表）。

根据Keras的文档，循环层需要一个形状为(batch_size, timesteps, input_dim)的数组。假设每一列对应不同的日期（如果我错了请纠正我），这样应该可以工作:

dim(data_train[, 1:3]) # [1] 10  3X <- as.matrix(data_train[, 1:3]) # 转换为数组dim(X) # [1] 10  3dim(X) <- c(dim(X), 1)dim(X) # [1] 10  3  1# 同样对YY <- as.matrix(data_train[, 4:5])dim(Y) <- c(dim(Y), 1)

现在X和Y具有3个维度，你可以将它们输入到你的模型中:

history = model %>% fit(x = X, y = Y, epochs = 40, batch_size = 20)