编辑:根据PM 2Ring 的建议,SymPy 可能是解决方案。
其他答案:
我认为TensorFlow 在这种情况下可能会对您有所帮助。
不过,在这种情况下,编写方程式可能会略有不同。
import tensorflow as tf
# define your variables before hand.
# tf.placeholder(<type of element>, <dimensions>)
arr1 = tf.placeholder(tf.float32, [2,2])
arr2 = tf.placeholder(tf.float32, [2,2])
# if it doesn't have dimension no need to mention.
scalar = tf.placeholder(tf.float32)
# these are your arrays which contain those arbitrary values you mentioned
multiplied_array = arr1*scalar
added_array = arr1+arr2
adding_to_multiplied = multiplied_array+arr1
# to evaluate these into any number, you need a session
sess = tf.Session()
# just provide values of arr1, scalar, arr2 to evaluate the results
print(sess.run(multiplied_array,{arr1:[[1,2],[3,4]],scalar:22}))
print(sess.run(added_array,{arr1:[[1,2],[3,4]],arr2:[[5,6],[7,8]]}))
print(sess.run(adding_to_multiplied,{arr1:[[1,2],[3,4]],arr2:[[5,6],[7,8]],scalar:22}))