如何将结构文本文件转换为熊猫数据框

Question

I have a text file with the following format:

"Model_1": {
        "10000": {
            "KPI_1": [
                0,
                3.0,
                6.0
            ],
            "KPI_2": [
                0,
                3.0,
                6.0
            ],
        }
            ...
        "15000": {
            "KPI_3": [
                0,
                3.0,
                6.0
            ],
            "KPI_4": [
                0,
                3.0,
                6.0
            ],
            ...
        }
        "20000": {
            "KPI_5": [
                0,
                3.0,
                6.0
            ],
            "KPI_6": [
                0,
                3.0,
                6.0
            ],
            ...
        }

}
"Model_2": {
        "10000": {
            "KPI_7": [
                v1,
                v2,
                v3
            ],
            "KPI_8": [
                v1,
                v2,
                v3
            ],
            ...
        }
        "15000": {
            "KPI_9": [
                v1,
                v2,
                v3
            ],
            "KPI_10": [
                v1,
                v2,
                v3
            ],
            ...
        }            
        "20000": {
            "KPI_11": [
                v1,
                v2,
                v3
            ],
            "KPI_12": [
                v1,
                v2,
                v3
            ],
            ...
        }            

}

And I would like to get a pandas df with the following structure:

       Model_1                            Model_2                         ... Model_N
       10000       15000      20000       10000       15000       20000       10000       15000       20000
KPI_1 {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}
KPI_2 {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}
KPI_3 {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}
...
KPI_N {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}  {v1,v2,v3}

Answer 1

The data from the text file that you show us look much like a dictionary, only comma's seem missing at two places where I put the comment # I added a comma here :

Model_1: {
    "10000": {
        "KPI_1": [
            0,
            3.0,
            6.0
        ],
        "KPI_2": [
            0,
            3.0,
            6.0
        ],
    }, # I added a comma here
    "15000": {
        "KPI_3": [
            0,
            3.0,
            6.0
        ],
        "KPI_4": [
            0,
            3.0,
            6.0
        ],
    }, # I added a comma here
    "20000": {
        "KPI_5": [
            0,
            3.0,
            6.0
        ],
        "KPI_6": [
            0,
            3.0,
            6.0
        ],
    }
} 

If it really are dictionaries, you can convert them to DataFrames:

pd.DataFrame.from_dict(Model_1, orient='index')