Question

我正在从Rbloggers复制示例，但是train函数会导致错误：

py_call_impl（可调用，dots $ args，dots $ keywords）错误： ValueError：列dtype和SparseTensors dtype必须兼容。键：ALUMNUS_IND，列dtype：，张量dtype：

该示例的原始代码没有ALUMNUS_IND的“ dtype = tf $ int32”，但是会导致相同的错误消息。有没有办法强制使用int32或完成火车功能的解决方案是什么？

tensorflow软件包是1.9

library(readr)
library(dplyr)
library(tensorflow)
library(tfestimators)


donor_data <- read_csv("https://www.dropbox.com/s/ntd5tbhr7fxmrr4/DonorSampleDataCleaned.csv?raw=1")

my_mode <- function(x) {
  ux <- unique(x)
  ux[which.max(tabulate(match(x, ux)))]
}

donor_data <- donor_data %>% 
  mutate_if(is.numeric, 
            .funs = funs(
              ifelse(is.na(.), 
                     median(., na.rm = TRUE),
                     .))) %>%
  mutate_if(is.character, 
            .funs = funs(
              ifelse(is.na(.), 
                     my_mode(.),
                     .)))

predictor_cols <- c("MARITAL_STATUS", "GENDER", 
                    "ALUMNUS_IND", "PARENT_IND", 
                    "WEALTH_RATING", "PREF_ADDRESS_TYPE")

# Convert feature to factor
donor_data <- mutate_at(donor_data, 
                        .vars = predictor_cols, 
                        .funs = as.factor)



feature_cols <- feature_columns(
  column_indicator(
    column_categorical_with_vocabulary_list(
      "MARITAL_STATUS", 
      vocabulary_list = unique(donor_data$MARITAL_STATUS))), 
  column_indicator(
    column_categorical_with_vocabulary_list(
      "GENDER", 
      vocabulary_list = unique(donor_data$GENDER))), 
  column_indicator(
    column_categorical_with_vocabulary_list(
      "ALUMNUS_IND", 
      vocabulary_list = unique(donor_data$ALUMNUS_IND),
      dtype = tf$int32)), 
  column_indicator(
    column_categorical_with_vocabulary_list(
      "PARENT_IND", 
      vocabulary_list = unique(donor_data$PARENT_IND))), 
  column_indicator(
    column_categorical_with_vocabulary_list(
      "WEALTH_RATING", 
      vocabulary_list = unique(donor_data$WEALTH_RATING))), 
  column_indicator(
    column_categorical_with_vocabulary_list(
      "PREF_ADDRESS_TYPE", 
      vocabulary_list = unique(donor_data$PREF_ADDRESS_TYPE))), 
  column_numeric("AGE"))


row_indices <- sample(1:nrow(donor_data), 
                      size = 0.8 * nrow(donor_data))
donor_data_train <- donor_data[row_indices, ]
donor_data_test <- donor_data[-row_indices, ]

donor_pred_fn <- function(data) {
  input_fn(data, 
           features = c("AGE", "MARITAL_STATUS", 
                        "GENDER", "ALUMNUS_IND", 
                        "PARENT_IND", "WEALTH_RATING", 
                        "PREF_ADDRESS_TYPE"), 
           response = "DONOR_IND")
}  

  classifier <- dnn_classifier(
    feature_columns = feature_cols, 
    hidden_units = c(80, 40, 30), 
    n_classes = 2, 
    label_vocabulary = c("N", "Y"))

  train(classifier, 
        input_fn = donor_pred_fn(donor_data_train))

Answer 1

注释以下代码以将其修复-

# Convert feature to factor
donor_data <- mutate_at(donor_data,.vars = predictor_cols, 
.funs = as.factor)

Answer 2

RDRR给出了很好的意见。请将.funs = as.factor更改为.funs = as.character。

# Convert feature to character
donor_data <- mutate_at(donor_data, 
                        .vars = predictor_cols, 
                        .funs = as.character)

r tensorflow ValueError：列dtype和SparseTensors dtype必须兼容

2 个答案: