"""
1. 安装kaggle的api: pip install kaggle
2. 在myAcount里找到create new API token，然后下载这个token，名字叫kaggle.json。
3. 把他放到/home/your account/.kaggle/文件夹内，如果没有这个文件夹则创建一个，注意这是一个默认隐藏的文件夹。
4. 要在比赛Rule page click "I understand and Accept"
5. 在终端输入对应的下载代码，这个在相应的kaggle数据集页面上有，比如我下的这个是kaggle competitions download -c challenges-in-representation-learning-facial-expression-recognition-challenge
"""
## 10 lines code to do Kaggle using autogluon
# env preparation
conda create -y --force -n ag python=3.8 pip
conda activate ag
pip install "mxnet<2.0.0"
pip install autogluon


## use case 1 with Kaggle titanic
# download data from kaggle
pip install kaggle
kaggle c download titanic
unzip -o titanic.zip

# starting model training
from autogluon.tabular import TabularDataset, TabularPredictor
train_data = TabularDataset('train.csv')
id, label = 'PassengerId', 'Survived'
predictor = TabularPredictor(label=label).fit(
	train_data.drop(columns=[id]))

# prediction and test
import pandas as pd
test_data = TabularDataset('test.csv')
preds = predictor.predict(test_data.drop(columns=[id]))
submission = pd.DataFrame({id:test_data[id], label:preds})
submission.to_csv('submission.csv', index=False)

## use case 2 with California house pricing
from autogluon.tabular import TabularDataset, TabularPredictor
import numpy as np

train_data = TabularDataset('train.csv')
id, label = 'PassengerId', 'Survived'
# data preprocessing
large_val_cols = ['Lot', 'Total interior livable area', 'Tax assessed value', 'Annual tax amount', 'Listed Price', 'Last Sold Price']
for c in large_val_cols + [label]:
	train_data[c] = np.log(train_data[c]+1)
# model 1
predictor = TabularPredictor(label=label).fit(train_data.drop(columns=[id]))

# model 2
predictor = TabularPredictor(label=label).fit(
        train_data.drop(columns=[id]),
        hyperparameters='multimodal',
        num_stack_levels=1,
        num_bag_folds=5)