如何存储一个Python脚本中变量的计算值，以便在另一个Python脚本中反复使用

我有一个python（3.7.0）脚本文件，它接收17000多个文本文件，然后逐行处理每个文件，将所需的数据填充到数据框中。因此，代码如下所示是Script1.py文件：

def filePrpcesserFunc(file):
   data = open(file, 'r')
   for line in data.readlines():
      #Does some computing here to get var1, var2, var3
      #var1, var2, var3 are all integers
   return var1, var2, var3

if __name__ == '__main__':
   import glob2
   import pandas
   files = glob2.glob("*.txt")
   #There are more than 17,000 text files to be computed
   X = pandas.DataFrame()
   for file in files:
      rvar1, rvar2, rvar3 = fileProcesserFunc(file)
      X.loc[str(rvar1), str(rvar2)] = rvar3

现在，上面的python脚本运行了17000多个文件来生成pandas dataframe X，它只需要计算一次就可以获得数据（而且计算非常昂贵）。我需要在另一个脚本文件（比如）中使用这个数据框来创建一个机器学习模型，因此为了测试我的模型，我需要一次又一次地访问X，也就是说，我需要一次又一次地运行Script2.py。 有人能提出什么建议吗？我将非常感谢您。

您可以使用内置模块：

在

script1.py

中：

import pickle
with open('dataframe.txt', 'wb') as f:
    pickle.dump(X, f)

import pickle
X = pickle.load(open('dataframe.txt', 'rb'))

X.to_pickle("./df.pkl")

import pandas as pd
X = pd.read_pickle(¨./df.pkl¨)

在

script2.py

中：

import pickle
with open('dataframe.txt', 'wb') as f:
    pickle.dump(X, f)

import pickle
X = pickle.load(open('dataframe.txt', 'rb'))

X.to_pickle("./df.pkl")

import pandas as pd
X = pd.read_pickle(¨./df.pkl¨)

您还可以在数据帧上加载内置方法：

在

script1.py

中：

import pickle
with open('dataframe.txt', 'wb') as f:
    pickle.dump(X, f)

import pickle
X = pickle.load(open('dataframe.txt', 'rb'))

X.to_pickle("./df.pkl")

import pandas as pd
X = pd.read_pickle(¨./df.pkl¨)

在

script2.py

中：

import pickle
with open('dataframe.txt', 'wb') as f:
    pickle.dump(X, f)

import pickle
X = pickle.load(open('dataframe.txt', 'rb'))

X.to_pickle("./df.pkl")

import pandas as pd
X = pd.read_pickle(¨./df.pkl¨)

您可以使用内置模块：

在

script1.py

中：

import pickle
with open('dataframe.txt', 'wb') as f:
    pickle.dump(X, f)

import pickle
X = pickle.load(open('dataframe.txt', 'rb'))

X.to_pickle("./df.pkl")

import pandas as pd
X = pd.read_pickle(¨./df.pkl¨)

在

script2.py

中：

import pickle
with open('dataframe.txt', 'wb') as f:
    pickle.dump(X, f)

import pickle
X = pickle.load(open('dataframe.txt', 'rb'))

X.to_pickle("./df.pkl")

import pandas as pd
X = pd.read_pickle(¨./df.pkl¨)

您还可以在数据帧上加载内置方法：

在

script1.py

中：

import pickle
with open('dataframe.txt', 'wb') as f:
    pickle.dump(X, f)

import pickle
X = pickle.load(open('dataframe.txt', 'rb'))

X.to_pickle("./df.pkl")

import pandas as pd
X = pd.read_pickle(¨./df.pkl¨)

在

script2.py

中：

import pickle
with open('dataframe.txt', 'wb') as f:
    pickle.dump(X, f)

import pickle
X = pickle.load(open('dataframe.txt', 'rb'))

X.to_pickle("./df.pkl")

import pandas as pd
X = pd.read_pickle(¨./df.pkl¨)