Error on Categorical Column

ValueError: Column dtype and SparseTensors dtype must be compatible. key: thal, column dtype: <dtype: ‘string’>, tensor dtype: <dtype: ‘int32’>

my code :

# EXERCISE: Create a categorical vocabulary column out of the
# above mentioned categories with the key specified as 'thal'.
thal = tf.feature_column.categorical_column_with_vocabulary_list(key='thal', 
                                                                 vocabulary_list=['fixed', 'normal', 'reversible'])# YOUR CODE HERE

# EXERCISE: Create an indicator column out of the created categorical column.
thal_one_hot = tf.feature_column.indicator_column(thal)# YOUR CODE HERE
demo(thal_one_hot)

Solution : Remember to cast dataframe[‘thal’] into string