{
  "id": 55238,
  "title": "What do these two warnings mean?",
  "url": "/competitions/talkingdata-adtracking-fraud-detection/discussion/55238",
  "author_name": "",
  "post_date": "2018-04-24T03:33:14.716898Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>These two warnings worry me. Seems they indicate the categorical features I defined were not interpreted as categorical?</p>\n\n<p><strong>UserWarning: Using categorical_feature in Dataset.\n  warnings.warn('Using categorical_feature in Dataset.')</strong></p>\n\n<p><strong>UserWarning: categorical_feature in param dict is overrided.\n  warnings.warn('categorical_feature in param dict is overrided.')</strong></p>\n\n<p>Here are my params:</p>\n\n<p>params = {\n    'boosting_type': 'dart', \n    'drop_rate': 0.1, \n    'objective': 'binary',\n    'metric': 'auc',\n    'learning_rate': 0.1,\n    'num_leaves': 7, <br>\n    'max_depth': 3, <br>\n    'min_child_samples': 100, \n    'max_bin': 100, \n    'subsample': 0.7, <br>\n    'subsample_freq': 1, \n    'colsample_bytree': 0.9,\n    'min_child_weight': 0, \n    'subsample_for_bin': 200000,\n    'min_split_gain': 0,\n    'reg_alpha': 0,\n    'reg_lambda': 0,\n    'nthread': 8,\n    'verbose': 0,\n    'scale_pos_weight': 200\n}</p>",
  "messages": [
    {
      "id": "318553",
      "postDate": "04/24/2018 03:33:14",
      "content": "<p>These two warnings worry me. Seems they indicate the categorical features I defined were not interpreted as categorical?</p>\n\n<p><strong>UserWarning: Using categorical_feature in Dataset.\n  warnings.warn('Using categorical_feature in Dataset.')</strong></p>\n\n<p><strong>UserWarning: categorical_feature in param dict is overrided.\n  warnings.warn('categorical_feature in param dict is overrided.')</strong></p>\n\n<p>Here are my params:</p>\n\n<p>params = {\n    'boosting_type': 'dart', \n    'drop_rate': 0.1, \n    'objective': 'binary',\n    'metric': 'auc',\n    'learning_rate': 0.1,\n    'num_leaves': 7, <br>\n    'max_depth': 3, <br>\n    'min_child_samples': 100, \n    'max_bin': 100, \n    'subsample': 0.7, <br>\n    'subsample_freq': 1, \n    'colsample_bytree': 0.9,\n    'min_child_weight': 0, \n    'subsample_for_bin': 200000,\n    'min_split_gain': 0,\n    'reg_alpha': 0,\n    'reg_lambda': 0,\n    'nthread': 8,\n    'verbose': 0,\n    'scale_pos_weight': 200\n}</p>",
      "rawMarkdown": "These two warnings worry me. Seems they indicate the categorical features I defined were not interpreted as categorical?\n\n\n**UserWarning: Using categorical_feature in Dataset.\n  warnings.warn('Using categorical_feature in Dataset.')**\n\n\n\n**UserWarning: categorical_feature in param dict is overrided.\n  warnings.warn('categorical_feature in param dict is overrided.')**\n\n\nHere are my params:\n\nparams = {\n    'boosting_type': 'dart', \n    'drop_rate': 0.1, \n    'objective': 'binary',\n    'metric': 'auc',\n    'learning_rate': 0.1,\n    'num_leaves': 7,  \n    'max_depth': 3,  \n    'min_child_samples': 100, \n    'max_bin': 100, \n    'subsample': 0.7,  \n    'subsample_freq': 1, \n    'colsample_bytree': 0.9,\n    'min_child_weight': 0, \n    'subsample_for_bin': 200000,\n    'min_split_gain': 0,\n    'reg_alpha': 0,\n    'reg_lambda': 0,\n    'nthread': 8,\n    'verbose': 0,\n    'scale_pos_weight': 200\n}",
      "votes": null
    },
    {
      "id": "318999",
      "postDate": "04/25/2018 02:06:39",
      "content": "<p>Spent some time digging into the source code. Turns out nothing to worry about. Categorical features still correctly adopted by the model definition.</p>",
      "rawMarkdown": "Spent some time digging into the source code. Turns out nothing to worry about. Categorical features still correctly adopted by the model definition.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 318999,
      "author_name": "enfeizhan",
      "author_url": "",
      "post_date": "04/25/2018 02:06:39",
      "content": "<p>Spent some time digging into the source code. Turns out nothing to worry about. Categorical features still correctly adopted by the model definition.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "318553": "These two warnings worry me. Seems they indicate the categorical features I defined were not interpreted as categorical?\n\n\n**UserWarning: Using categorical_feature in Dataset.\n  warnings.warn('Using categorical_feature in Dataset.')**\n\n\n\n**UserWarning: categorical_feature in param dict is overrided.\n  warnings.warn('categorical_feature in param dict is overrided.')**\n\n\nHere are my params:\n\nparams = {\n    'boosting_type': 'dart', \n    'drop_rate': 0.1, \n    'objective': 'binary',\n    'metric': 'auc',\n    'learning_rate': 0.1,\n    'num_leaves': 7,  \n    'max_depth': 3,  \n    'min_child_samples': 100, \n    'max_bin': 100, \n    'subsample': 0.7,  \n    'subsample_freq': 1, \n    'colsample_bytree': 0.9,\n    'min_child_weight': 0, \n    'subsample_for_bin': 200000,\n    'min_split_gain': 0,\n    'reg_alpha': 0,\n    'reg_lambda': 0,\n    'nthread': 8,\n    'verbose': 0,\n    'scale_pos_weight': 200\n}",
    "318999": "Spent some time digging into the source code. Turns out nothing to worry about. Categorical features still correctly adopted by the model definition."
  },
  "source": "meta"
}