将嵌套的类似字典的 txt 文件读入 Pandas 数据帧

Question

Sort of a new python guy here and haven't had much success with the following.

I have a txt file with data formatted as follows:

{
  "$type" : "TableInstance",
  "$version" : 1,
  "Instance" : "InstanceName",
  "ColumnAliases" : [ "", "", ],
  "ColumnNames" : [ "keyName", "dateName"],
  "ColumnData" : [ {
    "type" : "ColumnData1",
    "Strings" : [key1, key2],]
  }, {
    "type" : "ColumnData2",
    "Strings" : [date1, date2]}]
}

That I would like to read into a dataframe such that it is formatted as:

[   keyName     dateName
    key1        date1
    key2        date1 ]

Is there a simple way to do this?

Answer 1

does this work for you?

dict = {
  "$type" : "TableInstance",
  "$version" : 1,
  "Instance" : "InstanceName",
  "ColumnAliases" : [ "", "", ],
  "ColumnNames" : [ "keyName", "dateName"],
  "ColumnData" : [ {
    "type" : "ColumnData1",
    "Strings" : ['key1', 'key2']
  }, {
    "type" : "ColumnData2",
    "Strings" : ['date1', 'date2']}]
}

df = pd.DataFrame({dict['ColumnNames'][0]:dict['ColumnData'][0]['Strings'], dict['ColumnNames'][1]:dict['ColumnData'][1]['Strings']})

Answer 2

It looks it that you stored the serialized python object in the file. Hence, you can deserialize the Python object by the help of pickle, then you can parse the object based on your requirements.

import pickle
import pandas as pd

filePath = 'test.txt'
obj = pd.read_pickle(filePath)
#obj = pickle.load(open(filePath, "rb"))
df = pd.DataFrame({obj['ColumnNames'][0]:obj['ColumnData'][0]['Strings'], obj['ColumnNames'][1]:obj['ColumnData'][1]['Strings']})