Tabsyn Sampling Fails When Dataset Contains Only Categorical Features

[kirilzilla]

Description:

I encountered an error during the sampling phase using the tabsyn method. The issue seems to be related to missing numerical features during preprocessing.

Steps to Reproduce:

1. Use the following insurance.json configuration:
   It has 25 category columns and no numerical columns, the first column is the target column 0 or 1.

   {
       "name": "insurance",
       "task_type": "binclass",
       "header": "infer",
       "column_names": [
           "GoodStudent", "Age", "SocioEcon", "RiskAversion", "VehicleYear", 
           "RuggedAuto", "MakeModel", "DrivQuality", "Mileage", "Antilock", 
           "DrivingSkill", "SeniorTrain", "ThisCarCost", "Theft", "CarValue", 
           "HomeBase", "AntiTheft", "PropCost", "OtherCarCost", "OtherCar", 
           "MedCost", "Cushioning", "Airbag", "ILiCost", "DrivHist"
       ],
       "num_col_idx": [],
       "cat_col_idx": [
           1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 
           11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 
           21, 22, 23, 24
       ],
       "target_col_idx": [0],
       "file_type": "csv",
       "data_path": "data/insurance/insurance.csv",
       "test_path": null
   }
   

2. Preprocess and train the model successfully using VAE and tabsyn.

Preprocessing output shows a numerical vector, although there are no numerical features.
Numerical (18000, 0)
Categorical (18000, 24)

• I think this might be a mistake. It should be NO numerical columns and 25 categorical columns instead auf 24.

3. Run the sampling command:
   python main.py --dataname insurance --method tabsyn --mode sample

During the sampling step, the following error is thrown:

ValueError: Found array with 0 feature(s) (shape=(18000, 0)) while a minimum of 1 is required by QuantileTransformer.

This is the processed JSON in tabsyn/data/insurance/info.json

{
 "name": "insurance",
 "task_type": "binclass",
 "header": "infer",
 "column_names": [
     "GoodStudent",
     "Age",
     "SocioEcon",
     "RiskAversion",
     "VehicleYear",
     "RuggedAuto",
     "MakeModel",
     "DrivQuality",
     "Mileage",
     "Antilock",
     "DrivingSkill",
     "SeniorTrain",
     "ThisCarCost",
     "Theft",
     "CarValue",
     "HomeBase",
     "AntiTheft",
     "PropCost",
     "OtherCarCost",
     "OtherCar",
     "MedCost",
     "Cushioning",
     "Airbag",
     "ILiCost",
     "DrivHist"
 ],
 "num_col_idx": [],
 "cat_col_idx": [
     1,
     2,
     3,
     4,
     5,
     6,
     7,
     8,
     9,
     10,
     11,
     12,
     13,
     14,
     15,
     16,
     17,
     18,
     19,
     20,
     21,
     22,
     23,
     24
 ],
 "target_col_idx": [
     0
 ],
 "file_type": "csv",
 "data_path": "data/insurance/insurance.csv",
 "test_path": null,
 "column_info": {
     "1": {},
     "type": "categorical",
     "categorizes": [
         0,
         1
     ],
     "2": {},
     "3": {},
     "4": {},
     "5": {},
     "6": {},
     "7": {},
     "8": {},
     "9": {},
     "10": {},
     "11": {},
     "12": {},
     "13": {},
     "14": {},
     "15": {},
     "16": {},
     "17": {},
     "18": {},
     "19": {},
     "20": {},
     "21": {},
     "22": {},
     "23": {},
     "24": {},
     "0": {}
 },
 "train_num": 18000,
 "test_num": 2000,
 "idx_mapping": {
     "0": 24,
     "1": 0,
     "2": 1,
     "3": 2,
     "4": 3,
     "5": 4,
     "6": 5,
     "7": 6,
     "8": 7,
     "9": 8,
     "10": 9,
     "11": 10,
     "12": 11,
     "13": 12,
     "14": 13,
     "15": 14,
     "16": 15,
     "17": 16,
     "18": 17,
     "19": 18,
     "20": 19,
     "21": 20,
     "22": 21,
     "23": 22,
     "24": 23
 },
 "inverse_idx_mapping": {
     "24": 0,
     "0": 1,
     "1": 2,
     "2": 3,
     "3": 4,
     "4": 5,
     "5": 6,
     "6": 7,
     "7": 8,
     "8": 9,
     "9": 10,
     "10": 11,
     "11": 12,
     "12": 13,
     "13": 14,
     "14": 15,
     "15": 16,
     "16": 17,
     "17": 18,
     "18": 19,
     "19": 20,
     "20": 21,
     "21": 22,
     "22": 23,
     "23": 24
 },
 "idx_name_mapping": {
     "0": "GoodStudent",
     "1": "Age",
     "2": "SocioEcon",
     "3": "RiskAversion",
     "4": "VehicleYear",
     "5": "RuggedAuto",
     "6": "MakeModel",
     "7": "DrivQuality",
     "8": "Mileage",
     "9": "Antilock",
     "10": "DrivingSkill",
     "11": "SeniorTrain",
     "12": "ThisCarCost",
     "13": "Theft",
     "14": "CarValue",
     "15": "HomeBase",
     "16": "AntiTheft",
     "17": "PropCost",
     "18": "OtherCarCost",
     "19": "OtherCar",
     "20": "MedCost",
     "21": "Cushioning",
     "22": "Airbag",
     "23": "ILiCost",
     "24": "DrivHist"
 },
 "metadata": {
     "columns": {
         "1": {
             "sdtype": "categorical"
         },
         "2": {
             "sdtype": "categorical"
         },
         "3": {
             "sdtype": "categorical"
         },
         "4": {
             "sdtype": "categorical"
         },
         "5": {
             "sdtype": "categorical"
         },
         "6": {
             "sdtype": "categorical"
         },
         "7": {
             "sdtype": "categorical"
         },
         "8": {
             "sdtype": "categorical"
         },
         "9": {
             "sdtype": "categorical"
         },
         "10": {
             "sdtype": "categorical"
         },
         "11": {
             "sdtype": "categorical"
         },
         "12": {
             "sdtype": "categorical"
         },
         "13": {
             "sdtype": "categorical"
         },
         "14": {
             "sdtype": "categorical"
         },
         "15": {
             "sdtype": "categorical"
         },
         "16": {
             "sdtype": "categorical"
         },
         "17": {
             "sdtype": "categorical"
         },
         "18": {
             "sdtype": "categorical"
         },
         "19": {
             "sdtype": "categorical"
         },
         "20": {
             "sdtype": "categorical"
         },
         "21": {
             "sdtype": "categorical"
         },
         "22": {
             "sdtype": "categorical"
         },
         "23": {
             "sdtype": "categorical"
         },
         "24": {
             "sdtype": "categorical"
         },
         "0": {
             "sdtype": "categorical"
         }
     }
 }
}

[hengruizhang98]

Tabsyn now only supports data having both continuous and categorical features. To adapt to datasets fo only categorical features, you can rewrite the code to let it ignore the continuous inputs.