You need to enable JavaScript to run this app.
优惠活动
大模型
产品
解决方案
定价
更多

误用LinearClassifier致TensorFlow线性回归模型Assertion failed错误求助

avatar
数据老陈
2026-8-24
汽车价格预测线性回归模型训练错误排查与解决

问题描述

使用TensorFlow 2.0构建汽车价格预测的线性回归模型，操作流程如下：

  • 导入包含汽车特征（ID、Price、Manufacturer等）的CSV数据集并转为Pandas DataFrame
  • 重命名部分列名，移除无关特征
  • 分离特征数据与预测目标Price（作为标签）
  • 定义类别特征与数值特征并构建feature_columns
  • 创建输入函数，初始化模型并执行训练与评估

运行时触发Assertion failed错误，错误信息显示Labels must be <= n_classes - 1，标签值（如19444、56509等）远大于1，无法完成训练。

原代码 

# IMPORTS
import os
import numpy as np
import tensorflow as tf
import pandas as pd

# IMPORT CSV FILES AS PANDASDB
dfTrain = pd.read_csv('carPriceTrain.csv')
dfEval = pd.read_csv('carPriceEval.csv')

# RENAME COLUMN NAMES
dfTrain = dfTrain.rename(columns={"Gear box type": "gearBoxType"})
dfEval = dfTrain.rename(columns={"Gear box type": "gearBoxType"})
dfTrain = dfTrain.rename(columns={"Leather interior": "leatherInterior"})
dfEval = dfTrain.rename(columns={"Leather interior": "leatherInterior"})

# REMOVE EXTRANEOUS CAR INFORMATION
dfTrain.pop('ID')
dfTrain.pop('Levy')
dfTrain.pop('Category')
dfTrain.pop('Fuel type')
dfTrain.pop('Drive wheels')
dfTrain.pop('leatherInterior')
dfTrain.pop('gearBoxType')
dfTrain.pop('Doors')
dfTrain.pop('Wheel')
dfTrain.pop('Color')
dfTrain.pop('Airbags')

dfEval.pop('ID')
dfEval.pop('Levy')
dfEval.pop('Category')
dfEval.pop('Fuel type')
dfEval.pop('Drive wheels')
dfEval.pop('leatherInterior')
dfEval.pop('gearBoxType')
dfEval.pop('Doors')
dfEval.pop('Wheel')
dfEval.pop('Color')
dfEval.pop('Airbags')

# POP PRICE (PREDICTING VAL)
yTrain = dfTrain.pop('Price')
yEval = dfEval.pop('Price')

# CREATE COLUMNS NAMES
CATEGORICAL_COLUMNS = ['Manufacturer', 'Model']
NUMERIC_COLUMNS = ['Prod. year', 'Mileage', 'Cylinders', 'Engine volume']

# FEATURE COLUMNS FOR CATERGORICAL
featureColumns = []
for featureName in CATEGORICAL_COLUMNS:
  vocabulary = dfTrain[featureName].unique()
  featureColumns.append(tf.feature_column.categorical_column_with_vocabulary_list(featureName, vocabulary))

# FEATURE COLUMNS FOR NUMERIC
for featureName in NUMERIC_COLUMNS:
    featureColumns.append(tf.feature_column.numeric_column(featureName, dtype=tf.float32))

# CREATING INPUT FN
def makeInputFN(dfData, dfLabel, nEpochs=10, shuffle=True, batchSize=32):
  def inputFN():
    ds = tf.data.Dataset.from_tensor_slices((dict(dfData), dfLabel))
    if shuffle:
      ds = ds.shuffle(1000)
    ds = ds.batch(batchSize).repeat(nEpochs)
    return ds
  return inputFN

trainInputFN = makeInputFN(dfTrain, yTrain)
evalInputFN = makeInputFN(dfEval, yEval, nEpochs=1, shuffle=False)

linearEst = tf.estimator.LinearClassifier(feature_columns=featureColumns)

linearEst.train(trainInputFN)
result = linearEst.evaluate(evalInputFN)

错误信息 

WARNING:tensorflow:Using temporary folder as model directory: /var/folders/kq/823s5gqs0ds7wqckr9sxfcnm0000gn/T/tmp22bryhkx
WARNING:tensorflow:From /Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow/python/training/training_util.py:396: Variable.initialized_value (from tensorflow.python.ops.variables) is deprecated and will be removed in a future version.
Instructions for updating:
Use Variable.read_value. Variables in 2.X are initialized automatically both in eager and graph (inside tf.defun) contexts.
WARNING:tensorflow:From /Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/keras/optimizers/optimizer_v2/ftrl.py:153: calling Constant.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.
Instructions for updating:
Call initializer instance with the dtype argument instead of passing it to the constructor
2022-07-30 15:22:47.534048: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2022-07-30 15:22:47.549419: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:354] MLIR V1 optimization pass is not enabled
Traceback (most recent call last):
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow/python/client/session.py", line 1377, in _do_call
    return fn(*args)
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow/python/client/session.py", line 1360, in _run_fn
    return self._call_tf_sessionrun(options, feed_dict, fetch_list,
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow/python/client/session.py", line 1453, in _call_tf_sessionrun
    return tf_session.TF_SessionRun_wrapper(self._session, options, feed_dict,
tensorflow.python.framework.errors_impl.InvalidArgumentError: assertion failed: [Labels must be <= n_classes - 1] [Condition x <= y did not hold element-wise:] [x (head/losses/Cast:0) = ] [[19444][56509][8781]...] [y (head/losses/check_label_range/Const:0) = ] [1]
         [[{{function_node head_losses_check_label_range_assert_less_equal_Assert_AssertGuard_false_667}}{{node Assert}}]]

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/Users/anirud/CarPricePredictor/main.py", line 75, in <module>
    linearEst.train(trainInputFN)
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow_estimator/python/estimator/estimator.py", line 360, in train
    loss = self._train_model(input_fn, hooks, saving_listeners)
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow_estimator/python/estimator/estimator.py", line 1186, in _train_model
    return self._train_model_default(input_fn, hooks, saving_listeners)
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow_estimator/python/estimator/estimator.py", line 1217, in _train_model_default
    return self._train_with_estimator_spec(estimator_spec, worker_hooks,
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow_estimator/python/estimator/estimator.py", line 1533, in _train_with_estimator_spec
    _, loss = mon_sess.run([estimator_spec.train_op, estimator_spec.loss])
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow/python/training/monitored_session.py", line 782, in run
    return self._sess.run(
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow/python/training/monitored_session.py", line 1311, in run
    return self._sess.run(
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow/python/training/monitored_session.py", line 1416, in run
    raise six.reraise(*original_exc_info)
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/six.py", line 719, in reraise
    raise value
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow/python/training/monitored_session.py", line 1401, in run
    return self._sess.run(*args, **kwargs)
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow/python/training/monitored_session.py", line 1469, in run
    outputs = _WrappedSession.run(
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow/python/training/monitored_session.py", line 1232, in run
    return self._sess.run(*args, **kwargs)
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow/python/client/session.py", line 967, in run
    result = self._run(None, fetches, feed_dict, options_ptr,
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow/python/client/session.py", line 1190, in _run
    results = self._do_run(handle, final_targets, final_fetches,
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow/python/client/session.py", line 1370, in _do_run
    return self._do_call(_run_fn, feeds, fetches, targets, options,
  File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/tensorflow/python/client/session.py", line 1396, in _do_call
    raise type(e)(node_def, op, message)  # pylint: disable=no-value-for-parameter
tensorflow.python.framework.errors_impl.InvalidArgumentError: Graph execution error:

assertion failed: [Labels must be <= n_classes - 1] [Condition x <= y did not hold element-wise:] [x (head/losses/Cast:0) = ] [[19444][56509][8781]...] [y (head/losses/check_label_range/Const:0) = ] [1]
         [[{{node Assert}}]]

错误原因

  1. 模型类型错误：核心问题是使用了tf.estimator.LinearClassifier（线性分类器）处理回归任务。分类器默认是二分类模式（n_classes=2），强制要求标签值为0或1，但你的标签是连续的汽车价格，远超出分类器的标签范围，触发断言错误。
  2. 数据泄漏问题：代码中重命名和清理dfEval时，错误地将dfTrain赋值给dfEval（如dfEval = dfTrain.rename(...)），导致训练集和测试集完全相同，后续评估毫无意义，还会引入数据泄漏。

解决方法

  1. 替换为回归模型：将LinearClassifier替换为tf.estimator.LinearRegressor，这是TensorFlow专门用于线性回归任务的估算器，不需要标签满足分类范围要求。
  2. 修正数据处理逻辑：确保dfEval的操作基于自身，而非复制dfTrain的结果。
  3. 可选优化：对数值特征做归一化/标准化处理，回归模型对特征尺度敏感，归一化后能提升模型收敛速度和预测精度。

修正后的代码 

# IMPORTS
import os
import numpy as np
import tensorflow as tf
import pandas as pd

# IMPORT CSV FILES AS PANDASDB
dfTrain = pd.read_csv('carPriceTrain.csv')
dfEval = pd.read_csv('carPriceEval.csv')

# RENAME COLUMN NAMES - 修正dfEval的赋值逻辑
dfTrain = dfTrain.rename(columns={"Gear box type": "gearBoxType", "Leather interior": "leatherInterior"})
dfEval = dfEval.rename(columns={"Gear box type": "gearBoxType", "Leather interior": "leatherInterior"})

# REMOVE EXTRANEOUS CAR INFORMATION - 分别处理训练集和测试集
drop_cols = ['ID', 'Levy', 'Category', 'Fuel type', 'Drive wheels', 'leatherInterior', 'gearBoxType', 'Doors', 'Wheel', 'Color', 'Airbags']
dfTrain = dfTrain.drop(columns=drop_cols)
dfEval = dfEval.drop(columns=drop_cols)

# POP PRICE (PREDICTING VAL)
yTrain = dfTrain.pop('Price')
yEval = dfEval.pop('Price')

# CREATE COLUMNS NAMES
CATEGORICAL_COLUMNS = ['Manufacturer', 'Model']
NUMERIC_COLUMNS = ['Prod. year', 'Mileage', 'Cylinders', 'Engine volume']

# FEATURE COLUMNS FOR CATERGORICAL
featureColumns = []
for featureName in CATEGORICAL_COLUMNS:
  vocabulary = dfTrain[featureName].unique()
  # 分类特征需要转为indicator列才能输入回归模型
  cat_col = tf.feature_column.categorical_column_with_vocabulary_list(featureName, vocabulary)
  featureColumns.append(tf.feature_column.indicator_column(cat_col))

# FEATURE COLUMNS FOR NUMERIC - 添加归一化处理
def normalize_numeric_column(col_name):
  return tf.feature_column.numeric_column(
      col_name,
      dtype=tf.float32,
      normalizer_fn=lambda x: (x - dfTrain[col_name].mean()) / dfTrain[col_name].std()
  )

for featureName in NUMERIC_COLUMNS:
    featureColumns.append(normalize_numeric_column(featureName))

# CREATING INPUT FN
def makeInputFN(dfData, dfLabel, nEpochs=10, shuffle=True, batchSize=32):
  def inputFN():
    ds = tf.data.Dataset.from_tensor_slices((dict(dfData), dfLabel))
    if shuffle:
      ds = ds.shuffle(1000)
    ds = ds.batch(batchSize).repeat(nEpochs)
    return ds
  return inputFN

trainInputFN = makeInputFN(dfTrain, yTrain)
evalInputFN = makeInputFN(dfEval, yEval, nEpochs=1, shuffle=False)

# 替换为线性回归器
linearEst = tf.estimator.LinearRegressor(feature_columns=featureColumns)

linearEst.train(trainInputFN)
result = linearEst.evaluate(evalInputFN)

print("评估结果:", result)

关键修改说明

  • 将LinearClassifier改为LinearRegressor，适配回归任务。
  • 修正dfEval的数据处理逻辑，避免数据泄漏。
  • 分类特征通过indicator_column转为独热编码，符合回归模型的输入要求。
  • 对数值特征添加归一化处理，提升模型性能。

内容的提问来源于stack exchange，提问作者AnirudL

相关产品推荐
方舟 Agent Plan

超全模态模型 × Harness 升级，最新支持 Deepseek-V4.1-Flash、GLM-5.3 系列、Doubao-Seedream-5.0-pro、Kimi-K3 (部分)， 限时 9.9 元起

产品介绍相关文档
相关产品推荐
方舟 Agent Plan

超全模态模型 × Harness 升级，最新支持 Deepseek-V4.1-Flash、GLM-5.3 系列、Doubao-Seedream-5.0-pro、Kimi-K3 (部分)， 限时 9.9 元起

产品介绍相关文档
火山引擎 最新活动
全面开放
Doubao-Seedance-2.5全新模型升级

模型再升级,30秒超长叙事, 模态参考扩容

HOT
方舟Codingplan

模型自由，工具不限，最新支持 Deepseek-V4 系列、GLM-5.3 系列

HOT
方舟Agent plan

超全模态模型 × Harness 升级，最新支持 Deepseek-V4.1-Flash、GLM-5.3 系列

HOT
新用户特惠专场

大模型19元起，Al应用9.9元畅享，新人首购爆款尽享优惠

最近更新时间：2026.08.24 16:48:35