kernel_id int64 24.2k 23.3M | prompt stringlengths 8 1.85M | completetion stringlengths 1 182k | comp_name stringlengths 5 57 |
|---|---|---|---|
17,910,828 | train_data = df_train.drop(['id','keyword','location'], axis=1)
train_data.to_csv('cleaned_train.csv', index=False)
test_data = df_test.drop(['keyword','location'], axis=1)
test_data.to_csv('cleaned_test.csv', index=False )<load_from_csv> | preds = clf.predict_proba(apps_all_test.drop('SK_ID_CURR', axis=1)) [:, 1 ]
apps_all_test['TARGET'] = preds
apps_all_test[['SK_ID_CURR', 'TARGET']].to_csv('prev_baseline_03.csv', index=False ) | Home Credit Default Risk |
17,544,955 | train_data = pd.read_csv('cleaned_train.csv')
len(train_data)
<set_options> | application_train = pd.read_csv('/kaggle/input/home-credit-default-risk/application_train.csv')
application_test = pd.read_csv('/kaggle/input/home-credit-default-risk/application_test.csv')
| Home Credit Default Risk |
17,544,955 | SEED = 1234
random.seed(SEED)
np.random.seed(SEED)
torch.manual_seed(SEED)
torch.backends.cudnn.deterministic = True<load_pretrained> | print("Dimension of application_train :", application_train.shape)
print("결측치가 있는 컬럼 수 :",(application_train.isnull().sum() !=0 ).sum())
application_train.head() | Home Credit Default Risk |
17,544,955 | tokenizer = BertTokenizer.from_pretrained('bert-base-uncased' )<string_transform> | print("Dimension :", application_train.dropna(axis=0 ).shape)
print("결측치가 있는 컬럼 수 :",(application_train.dropna(axis=0 ).isnull().sum() !=0 ).sum())
application_train.dropna(axis=0 ) | Home Credit Default Risk |
17,544,955 | init_token = tokenizer.cls_token
eos_token = tokenizer.sep_token
pad_token = tokenizer.pad_token
unk_token = tokenizer.unk_token<categorify> | column_list = []
for name in column_series.keys() :
if(column_series[name]>100000):
column_list.append(name)
print(column_list, len(column_list)) | Home Credit Default Risk |
17,544,955 | init_token_idx = tokenizer.convert_tokens_to_ids(init_token)
eos_token_idx = tokenizer.convert_tokens_to_ids(eos_token)
pad_token_idx = tokenizer.convert_tokens_to_ids(pad_token)
unk_token_idx = tokenizer.convert_tokens_to_ids(unk_token)
print(init_token_idx, eos_token_idx, pad_token_idx, unk_token_idx )<define_var... | def show_hist_by_target(df, columns):
cond_1 =(df['TARGET'] == 1)
cond_0 =(df['TARGET'] == 0)
for column in columns:
fig, ax = plt.subplots(figsize=(12, 4), nrows=1, ncols=2, squeeze=False)
if(type(df[column][0])is str):
df_temp = df[["TARGET",column]].value_counts().astype(float)
idx_temp = df_temp.reset_index(nam... | Home Credit Default Risk |
17,544,955 | max_input_length = tokenizer.max_model_input_sizes['bert-base-uncased']
print(max_input_length )<string_transform> | abs(cor["TARGET"] ).sort_values() | Home Credit Default Risk |
17,544,955 | def tokenize_and_cut(sentence):
tokens = tokenizer.tokenize(sentence)
tokens = tokens[:max_input_length-2]
return tokens<data_type_conversions> | application_train.dtypes.value_counts() | Home Credit Default Risk |
17,544,955 | TEXT = data.Field(batch_first = True,
use_vocab = False,
tokenize = tokenize_and_cut,
preprocessing = tokenizer.convert_tokens_to_ids,
init_token = init_token_idx,
eos_token = eos_token_idx,
pad_token = pad_token_idx,
unk_token = unk_token_idx)
LABEL = data.LabelField(dtype = torch.float )<split> | application_train["FONDKAPREMONT_MODE"] | Home Credit Default Risk |
17,544,955 | fields = [('text', TEXT),('target', LABEL)]
datasets = torchtext.legacy.data.TabularDataset(
path='cleaned_train.csv',format='csv',skip_header=True,fields=fields)
train_data, test_data = datasets.split(split_ratio=[0.95, 0.05])
train_data, valid_data = train_data.split(random_state = random.seed(SEED))<feature_engin... | le = LabelEncoder()
le_count = 0
for col in application_train:
if application_train[col].dtype == 'object':
if len(list(application_train[col].unique())) >= 2:
le.fit(application_train[col])
application_train[col] = le.transform(application_train[col])
application_test[col] = le.transform(application_test[col])
le_c... | Home Credit Default Risk |
17,544,955 | LABEL.build_vocab(train_data )<split> | application_train["FONDKAPREMONT_MODE"] | Home Credit Default Risk |
17,544,955 | BATCH_SIZE = 128
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
print(device)
train_iterator, valid_iterator, test_iterator = data.BucketIterator.splits(
(train_data, valid_data, test_data),
batch_size = BATCH_SIZE,
device = device )<split> | application_train.select_dtypes('object' ).apply(pd.Series.nunique, axis = 0 ) | Home Credit Default Risk |
17,544,955 | train_data, valid_data = train_data.split(
split_ratio=[0.85, 0.15],
random_state=random.seed(123))
print('Num Train: {}'.format(len(train_data)))
print('Num Validation: {}'.format(len(valid_data)) )<load_pretrained> | rel_list = []
for rel_column in rel.index:
if(rel[rel_column]<0.03):
rel_list.append(rel_column)
print(rel_column ) | Home Credit Default Risk |
17,544,955 | bert = BertModel.from_pretrained('bert-base-uncased' )<import_modules> | rel_list.remove('SK_ID_CURR' ) | Home Credit Default Risk |
17,544,955 | class BERTGRUDisaster(nn.Module):
def __init__(self,
bert,
hidden_dim,
output_dim,
n_layers,
bidirectional,
dropout):
super().__init__()
self.bert = bert
embedding_dim = bert.config.to_dict() ['hidden_size']
self.rnn = nn.GRU(embedding_dim,
hidden_dim,
num_layers = n_layers,
bidirectional = bidirectional,
batch_first =... | column_list.remove("EXT_SOURCE_1")
app_train = application_train | Home Credit Default Risk |
17,544,955 | HIDDEN_DIM = 256
OUTPUT_DIM = 1
N_LAYERS = 2
BIDIRECTIONAL = True
DROPOUT = 0.25
model = BERTGRUDisaster(bert,
HIDDEN_DIM,
OUTPUT_DIM,
N_LAYERS,
BIDIRECTIONAL,
DROPOUT )<find_best_params> | print("Dimension of application_test :", application_test.shape)
print("결측치가 있는 컬럼 수 :",(application_test.isnull().sum() !=0 ).sum())
application_test.head() | Home Credit Default Risk |
17,544,955 | for name, param in model.named_parameters() :
if name.startswith('bert'):
param.requires_grad = False<choose_model_class> | app_test = application_test | Home Credit Default Risk |
17,544,955 | optimizer = optim.Adam(model.parameters())
criterion = nn.BCEWithLogitsLoss()<find_best_params> | def data_processing(out, data):
out['APPS_EXT_SOURCE_MEAN'] = data[['EXT_SOURCE_1', 'EXT_SOURCE_2', 'EXT_SOURCE_3']].mean(axis=1)
out['APPS_EXT_SOURCE_STD'] = data[['EXT_SOURCE_1', 'EXT_SOURCE_2', 'EXT_SOURCE_3']].std(axis=1)
out['APPS_EXT_SOURCE_STD'] = out['APPS_EXT_SOURCE_STD'].fillna(out['APPS_EXT_SOURCE_STD'].me... | Home Credit Default Risk |
17,544,955 | model = model.to(device)
criterion = criterion.to(device )<compute_test_metric> | app_train = data_processing(app_train, application_train)
app_test = data_processing(app_test, application_test)
app_train.shape, app_test.shape | Home Credit Default Risk |
17,544,955 | def binary_accuracy(preds, y):
rounded_preds = torch.round(torch.sigmoid(preds))
correct =(rounded_preds == y ).float()
acc = correct.sum() / len(correct)
return acc<train_on_grid> | prev_app = pd.read_csv('.. /input/home-credit-default-risk/previous_application.csv')
print(prev_app.shape, app_train.shape ) | Home Credit Default Risk |
17,544,955 | def train(model, iterator, optimizer, criterion):
epoch_loss = 0
epoch_acc = 0
model.train()
for batch in iterator:
optimizer.zero_grad()
predictions = model(batch.text ).squeeze(1)
loss = criterion(predictions, batch.target)
acc = binary_accuracy(predictions, batch.target)
loss.backward()
optimizer.step()
epoch_los... | prev_app['PREV_CREDIT_DIFF'] = prev_app['AMT_APPLICATION'] - prev_app['AMT_CREDIT']
prev_app['PREV_GOODS_DIFF'] = prev_app['AMT_APPLICATION'] - prev_app['AMT_GOODS_PRICE']
prev_app['PREV_CREDIT_APPL_RATIO'] = prev_app['AMT_CREDIT']/prev_app['AMT_APPLICATION']
prev_app['PREV_ANNUITY_APPL_RATIO'] = prev_app['AMT_ANNUITY'... | Home Credit Default Risk |
17,544,955 | def epoch_time(start_time, end_time):
elapsed_time = end_time - start_time
elapsed_mins = int(elapsed_time / 60)
elapsed_secs = int(elapsed_time -(elapsed_mins * 60))
return elapsed_mins, elapsed_secs<compute_test_metric> | prev_app['DAYS_FIRST_DRAWING'].replace(365243, np.nan, inplace=True)
prev_app['DAYS_FIRST_DUE'].replace(365243, np.nan, inplace= True)
prev_app['DAYS_LAST_DUE_1ST_VERSION'].replace(365243, np.nan, inplace= True)
prev_app['DAYS_LAST_DUE'].replace(365243, np.nan, inplace= True)
prev_app['DAYS_TERMINATION'].replace(36... | Home Credit Default Risk |
17,544,955 | def binary_accuracy(preds, y):
rounded_preds = torch.round(torch.sigmoid(preds))
correct =(rounded_preds == y ).float()
acc = correct.sum() / len(correct)
return acc<train_model> | agg_dict = {
'SK_ID_CURR':['count'],
'AMT_CREDIT':['mean', 'max', 'sum'],
'AMT_ANNUITY':['mean', 'max', 'sum'],
'AMT_APPLICATION':['mean', 'max', 'sum'],
'AMT_DOWN_PAYMENT':['mean', 'max', 'sum'],
'AMT_GOODS_PRICE':['mean', 'max', 'sum'],
'RATE_DOWN_PAYMENT': ['min', 'max', 'mean'],
'DAYS_DECISION': ['min', 'max', 'mea... | Home Credit Default Risk |
17,544,955 | N_EPOCHS = 5
best_valid_loss = float('inf')
for epoch in range(N_EPOCHS):
start_time = time.time()
train_loss, train_acc = train(model, train_iterator, optimizer, criterion)
end_time = time.time()
epoch_mins, epoch_secs = epoch_time(start_time, end_time)
print(f'Epoch: {epoch+1:02} | Epoch Time: {epoch_mins}m {epoch... | prev_app_merge = app_train.merge(prev_amt_agg, on='SK_ID_CURR', how='left', indicator=True)
prev_app_merge = prev_app_merge.drop(columns=['_merge'])
prev_app_merge.shape | Home Credit Default Risk |
17,544,955 | torch.save(model.state_dict() , 'disaster-model.pt' )<predict_on_test> | prev_app['NAME_CONTRACT_STATUS'].value_counts() | Home Credit Default Risk |
17,544,955 | def predict_disaster(model, tokenizer, sentence):
model.eval()
tokens = tokenizer.tokenize(sentence)
tokens = tokens[:max_input_length-2]
indexed = [init_token_idx] + tokenizer.convert_tokens_to_ids(tokens)+ [eos_token_idx]
tensor = torch.LongTensor(indexed ).to(device)
tensor = tensor.unsqueeze(0)
prediction = torc... | cond_refused =(prev_app['NAME_CONTRACT_STATUS'] == 'Refused')
cond_approved =(prev_app['NAME_CONTRACT_STATUS'] == 'Approved')
prev_refused = prev_app[cond_refused]
prev_approved = prev_app[cond_approved]
prev_refused.shape, prev_approved.shape, prev_app.shape | Home Credit Default Risk |
17,544,955 | predict_disaster(model, tokenizer, "Our Deeds are the Reason of this<load_from_csv> | prev_refused = prev_refused.groupby('SK_ID_CURR')
prev_approved = prev_approved.groupby('SK_ID_CURR' ) | Home Credit Default Risk |
17,544,955 | test_data = pd.read_csv('cleaned_test.csv')
test_data.head(10 )<count_missing_values> | prev_refused = prev_refused['NAME_CONTRACT_TYPE'].count()
prev_refused.name = "PRE_CONTRACT_REFUSED"
prev_approved = prev_approved['NAME_CONTRACT_TYPE'].count()
prev_approved.name = "PRE_CONTRACT_APPROVED" | Home Credit Default Risk |
17,544,955 | test_data = test_data.fillna('nan')
test_data.isna().sum()<categorify> | prev_app_merge = prev_app_merge.merge(prev_approved, on='SK_ID_CURR', how='left', indicator=False)
prev_app_merge = prev_app_merge.merge(prev_refused, on='SK_ID_CURR', how='left', indicator=False)
prev_app_merge['PRE_CONTRACT_APPROVED_RATE'] = prev_app_merge['PRE_CONTRACT_APPROVED'] /(prev_app_merge['PRE_CONTRACT_APP... | Home Credit Default Risk |
17,544,955 | submission_dict = {'id' : [], 'target' : []}
for data in test_data.iterrows() :
idx = data[1].id
text = data[1].text
target = predict_disaster(model, tokenizer, text)
target = 0 if target < 0.5 else 1
submission_dict['id'].append(idx)
submission_dict['target'].append(target)
<create_dataframe> | prev_app_merge = prev_app_merge.replace(float('NaN'),0)
prev_app_merge.head() | Home Credit Default Risk |
17,544,955 | sample_df = pd.DataFrame(submission_dict)
sample_df<save_to_csv> | bureau = pd.read_csv('.. /input/home-credit-default-risk/bureau.csv')
print("Size of bureau data", bureau.shape ) | Home Credit Default Risk |
17,544,955 | sample_df.to_csv('sample_submission_01.csv', index=False )<load_from_csv> | PAST_LOANS_PER_CUS = bureau[['SK_ID_CURR', 'DAYS_CREDIT']].groupby(by = ['SK_ID_CURR'])['DAYS_CREDIT'].count().reset_index().rename(index=str, columns={'DAYS_CREDIT': 'BUREAU_LOAN_COUNT'})
app_train_bureau = prev_app_merge.merge(PAST_LOANS_PER_CUS, on = ['SK_ID_CURR'], how = 'left')
print(app_train_bureau.shape)
app... | Home Credit Default Risk |
17,544,955 | x = pd.read_csv('sample_submission_01.csv')
x.head()
<set_options> | BUREAU_LOAN_TYPES = bureau[['SK_ID_CURR', 'CREDIT_TYPE']].groupby(by = ['SK_ID_CURR'])['CREDIT_TYPE'].nunique().reset_index().rename(index=str, columns={'CREDIT_TYPE': 'BUREAU_LOAN_TYPES'})
app_train_bureau = app_train_bureau.merge(BUREAU_LOAN_TYPES, on = ['SK_ID_CURR'], how = 'left' ).fillna(0)
print(app_train_burea... | Home Credit Default Risk |
17,544,955 | warnings.filterwarnings('ignore')
<load_from_csv> | app_train_bureau['AVERAGE_LOAN_TYPE'] = app_train_bureau['BUREAU_LOAN_COUNT']/app_train_bureau['BUREAU_LOAN_TYPES']
app_train_bureau = app_train_bureau.fillna(0)
print(app_train_bureau.shape)
app_train_bureau.head() | Home Credit Default Risk |
17,544,955 | train = pd.read_csv('.. /input/nlp-getting-started/train.csv', usecols=['id','text','target'])
test = pd.read_csv('.. /input/nlp-getting-started/test.csv', usecols=['id','text'])
sample = pd.read_csv('.. /input/nlp-getting-started/sample_submission.csv' )<categorify> | del app_train_bureau['BUREAU_LOAN_COUNT'], app_train_bureau['BUREAU_LOAN_TYPES']
app_train_bureau.head() | Home Credit Default Risk |
17,544,955 | %%time
def clean(tweet):
tweet = re.sub(r"\x89Û_", "", tweet)
tweet = re.sub(r"\x89ÛÒ", "", tweet)
tweet = re.sub(r"\x89ÛÓ", "", tweet)
tweet = re.sub(r"\x89ÛÏWhen", "When", tweet)
tweet = re.sub(r"\x89ÛÏ", "", tweet)
tweet = re.sub(r"China\x89Ûªs", "China's", tweet)
tweet = re.sub(r"let\x89Ûªs", "let's", tweet)
... | def f(x):
if x == 'Closed':
y = 0
else:
y = 1
return y
bureau_fe1 = bureau
bureau_fe1['CREDIT_ACTIVE_CLOSED'] = bureau_fe1.apply(lambda x: f(x.CREDIT_ACTIVE), axis = 1)
bureau_fe1.head() | Home Credit Default Risk |
17,544,955 | train['text'] = train['text'].apply(lambda s : clean(s))<filter> | grp = bureau_fe1.groupby(by = ['SK_ID_CURR'])['CREDIT_ACTIVE_CLOSED'].mean().reset_index().rename(index=str, columns={'CREDIT_ACTIVE_CLOSED':'ACTIVE_LOANS_PERCENTAGE'})
app_train_bureau = app_train_bureau.merge(grp, on = ['SK_ID_CURR'], how = 'left')
del bureau_fe1['CREDIT_ACTIVE_CLOSED']
print(bureau_fe1.shape)
bur... | Home Credit Default Risk |
17,544,955 | train[train.target == 0]<create_dataframe> | app_train_bureau = app_train_bureau.fillna(0)
app_train_bureau.head() | Home Credit Default Risk |
17,544,955 | train_cleaned_df = train.copy()<load_pretrained> | app_train_bureau['BUREAU_ENDDATE_FACT_DIFF'] = bureau['DAYS_CREDIT_ENDDATE'] - bureau['DAYS_ENDDATE_FACT']
app_train_bureau['BUREAU_CREDIT_FACT_DIFF'] = bureau['DAYS_CREDIT'] - bureau['DAYS_ENDDATE_FACT']
app_train_bureau['BUREAU_CREDIT_ENDDATE_DIFF'] = bureau['DAYS_CREDIT'] - bureau['DAYS_CREDIT_ENDDATE']
app_train_bu... | Home Credit Default Risk |
17,544,955 | tokenizer = AutoTokenizer.from_pretrained('bert-large-uncased')
bert = TFBertModel.from_pretrained('bert-large-uncased' )<string_transform> | app_train = app_train_bureau | Home Credit Default Risk |
17,544,955 | tokenizer('Shine on you crazy diamond.' )<string_transform> | ftr_app = app_train.drop(columns=['SK_ID_CURR','TARGET'])
target_app = app_train['TARGET']
train_x, valid_x, train_y, valid_y = train_test_split(ftr_app, target_app, test_size=0.3, random_state=2020)
train_x.shape, valid_x.shape | Home Credit Default Risk |
17,544,955 | print("max len of tweets",max([len(x.split())for x in train.text]))<string_transform> | def lgb_cv(num_leaves, learning_rate, n_estimators, subsample, colsample_bytree, reg_alpha, reg_lambda, x_data=None, y_data=None, n_splits=5, output='score'):
score = 0
kf = KFold(n_splits=n_splits)
models = []
for train_index, valid_index in kf.split(x_data):
x_train, y_train = x_data.reindex([train_index]), y_data.r... | Home Credit Default Risk |
17,544,955 | x_train = tokenizer(
text=train.text.tolist() ,
add_special_tokens=True,
max_length=73,
truncation=True,
padding=True,
return_tensors='tf',
return_token_type_ids = False,
return_attention_mask = True,
verbose = True)
<count_values> | func_fixed = partial(lgb_cv, x_data=train_x, y_data=train_y, n_splits=5, output='score')
lgbBO = BayesianOptimization(
func_fixed,
{
'num_leaves':(16, 1024),
'learning_rate':(0.0001, 0.1),
'n_estimators':(16, 1024),
'subsample':(0, 1),
'colsample_bytree':(0, 1),
'reg_alpha':(0, 10),
'reg_lambda':(0, 50),
},
random_st... | Home Credit Default Risk |
17,544,955 | train.target.value_counts()<choose_model_class> | clf = LGBMClassifier(
n_estimators=int(lgbBO.max['params']['n_estimators']),
learning_rate=lgbBO.max['params']['learning_rate'],
num_leaves=int(lgbBO.max['params']['num_leaves']),
subsample=lgbBO.max['params']['subsample'],
max_depth=16,
reg_alpha=lgbBO.max['params']['reg_alpha'],
reg_lambda=lgbBO.max['params']['reg_l... | Home Credit Default Risk |
17,544,955 | <choose_model_class><EOS> | test_merge = app_test.merge(prev_amt_agg, on='SK_ID_CURR', how='left', indicator=False)
test_merge = test_merge.merge(prev_approved, on='SK_ID_CURR', how='left', indicator=False)
test_merge = test_merge.merge(prev_refused, on='SK_ID_CURR', how='left', indicator=False)
test_merge['PRE_CONTRACT_APPROVED_RATE'] = test_... | Home Credit Default Risk |
15,886,745 | <SOS> metric: AUC Kaggle data source: home-credit-default-risk<train_model> | %matplotlib inline | Home Credit Default Risk |
15,886,745 | train_history = model.fit(
x ={'input_ids':x_train['input_ids'],'attention_mask':x_train['attention_mask']} ,
y = y_train, epochs=12, batch_size=32
)<string_transform> | import os, sys | Home Credit Default Risk |
15,886,745 | x_test = tokenizer(
text=test.text.tolist() ,
add_special_tokens=True,
max_length=73,
truncation=True,
padding=True,
return_tensors='tf',
return_token_type_ids = False,
return_attention_mask = True,
verbose = True)
<predict_on_test> | default_dir = ".. /input/home-credit-default-risk/" | Home Credit Default Risk |
15,886,745 | predicted = model.predict({'input_ids':x_test['input_ids'],'attention_mask':x_test['attention_mask']} )<prepare_x_and_y> | def get_balance_data() :
pos_dtype = {
'SK_ID_PREV':np.uint32, 'SK_ID_CURR':np.uint32, 'MONTHS_BALANCE':np.int32, 'SK_DPD':np.int32,
'SK_DPD_DEF':np.int32, 'CNT_INSTALMENT':np.float32,'CNT_INSTALMENT_FUTURE':np.float32
}
install_dtype = {
'SK_ID_PREV':np.uint32, 'SK_ID_CURR':np.uint32, 'NUM_INSTALMENT_NUMBER':np.int32,... | Home Credit Default Risk |
15,886,745 | y_predicted = np.where(predicted>0.5,1,0 )<prepare_output> | from sklearn.model_selection import train_test_split
from lightgbm import LGBMClassifier | Home Credit Default Risk |
15,886,745 | y_predictedd = y_predicted.reshape(( 1,3263)) [0]<feature_engineering> | def get_apps_processed(apps):
apps['APPS_EXT_SOURCE_MEAN'] = apps[['EXT_SOURCE_1', 'EXT_SOURCE_2', 'EXT_SOURCE_3']].mean(axis = 1)
apps['APPS_EXT_SOURCE_STD'] = apps[['EXT_SOURCE_1', 'EXT_SOURCE_2', 'EXT_SOURCE_3']].std(axis=1)
apps['APPS_EXT_SOURCE_STD'] = apps['APPS_EXT_SOURCE_STD'].fillna(apps['APPS_EXT_SOURCE_S... | Home Credit Default Risk |
15,886,745 | sample['id'] = test.id
sample['target'] = y_predictedd<save_to_csv> | def get_pos_bal_agg(pos_bal):
cond_over_0 = pos_bal['SK_DPD'] > 0
cond_100 =(pos_bal['SK_DPD'] < 100)&(pos_bal['SK_DPD'] > 0)
cond_over_100 =(pos_bal['SK_DPD'] >= 100)
pos_bal['POS_IS_DPD'] = pos_bal['SK_DPD'].apply(lambda x: 1 if x > 0 else 0)
pos_bal['POS_IS_DPD_UNDER_120'] = pos_bal['SK_DPD'].apply(lambda x:1 if(... | Home Credit Default Risk |
15,886,745 | sample.to_csv('submission_a.csv',index = False )<load_from_csv> | def get_apps_all_with_all_agg(apps, prev, bureau, bureau_bal, pos_bal, install, card_bal):
apps_all = get_apps_processed(apps)
prev_agg = get_prev_agg(prev)
bureau_agg = get_bureau_agg(bureau, bureau_bal)
pos_bal_agg = get_pos_bal_agg(pos_bal)
install_agg = get_install_agg(install)
card_bal_agg = get_card_bal_ag... | Home Credit Default Risk |
15,886,745 | df = pd.read_csv("/kaggle/input/nlp-getting-started/train.csv", na_filter=False)
df.head()<categorify> | def get_dataset() :
app_train = pd.read_csv(os.path.join(default_dir,'application_train.csv'))
app_test = pd.read_csv(os.path.join(default_dir,'application_test.csv'))
apps = pd.concat([app_train, app_test])
prev = pd.read_csv(os.path.join(default_dir,'previous_application.csv'))
bureau = pd.read_csv(os.path.join(de... | Home Credit Default Risk |
15,886,745 | nlp = spacy.load("en_core_web_sm")
def preprocess(text):
doc = nlp(text)
token_semstop = [word for word in doc if not word.is_stop if not word.text == '
text = ' '.join(token.lower_ for token in token_semstop)
text = re.sub(r'(@\w+|https?:\S+)', '', text)
text = text.replace(r'&?', r'and')
text = re.sub(r'(>... | apps, prev, bureau, bureau_bal, pos_bal, install, card_bal = get_dataset() | Home Credit Default Risk |
15,886,745 | train_df['text'] = train_df['text'].apply(preprocess)
train_df.head()<load_from_csv> | apps_all = get_apps_all_with_all_agg(apps, prev, bureau, bureau_bal, pos_bal, install, card_bal)
apps_all = get_apps_all_encoded(apps_all)
apps_all_train, apps_all_test = get_apps_all_train_test(apps_all)
clf = train_apps_all(apps_all_train ) | Home Credit Default Risk |
15,886,745 | second_df = pd.read_csv("/kaggle/input/nlp-getting-started/test.csv", na_filter=False)
second_df.head()
test_df = second_df[['text']].copy()
test_df['text'] = test_df['text'].apply(preprocess)
<normalization> | output_dir = ".. /output/kaggle/working/"
preds = clf.predict_proba(apps_all_test.drop(['SK_ID_CURR'], axis=1)) [:, 1 ]
apps_all_test['TARGET'] = preds
apps_all_test[['SK_ID_CURR', 'TARGET']] | Home Credit Default Risk |
15,886,745 | vectorizer = TfidfVectorizer(use_idf=True, ngram_range=(1,2), preprocessor=preprocess)
tfidf_data = vectorizer.fit_transform(train_df['text'])
<find_best_model_class> | apps_all_test[['SK_ID_CURR', 'TARGET']].to_csv('submission.csv', index=False ) | Home Credit Default Risk |
15,886,745 | <load_from_csv><EOS> | from lightgbm import plot_importance | Home Credit Default Risk |
22,248,594 | <SOS> metric: AUC Kaggle data source: home-credit-default-risk<save_to_csv> | import numpy as np
import pandas as pd
import joblib
import gc | Home Credit Default Risk |
22,248,594 | submission['target'] = test_df['target']
submission.to_csv("sample_submission.csv", index=False)
submission.head()<install_modules> | test = pd.read_csv('.. /input/home-credit-default-risk/application_test.csv')
test.set_index(['SK_ID_CURR'], inplace=True)
test.shape | Home Credit Default Risk |
22,248,594 | !pip install --user catboost
<load_from_csv> | preprocessor = joblib.load('.. /input/wk6-default/wk6default_preprocessor.joblib')
LGBM_model = joblib.load('.. /input/wk6-default/wk6_LGBM_default_model.joblib' ) | Home Credit Default Risk |
22,248,594 | train = pd.read_csv('train.csv')
test = pd.read_csv('test.csv' )<count_missing_values> | bureau_bal = pd.read_csv('.. /input/home-credit-default-risk/bureau_balance.csv')
bureau = pd.read_csv('.. /input/home-credit-default-risk/bureau.csv')
bb = pd.merge(bureau, bureau_bal, on = 'SK_ID_BUREAU', how = 'left')
bb['REMAIN_CRED'] = bb['AMT_CREDIT_SUM'] - bb['AMT_CREDIT_SUM_DEBT'] - bb['AMT_CREDIT_SUM_LIMIT'... | Home Credit Default Risk |
22,248,594 | print(train.isnull().sum())
test.isnull().sum()<prepare_x_and_y> | cc_bal = pd.read_csv('.. /input/home-credit-default-risk/credit_card_balance.csv')
cc_bal['DRAW_RATIO'] = cc_bal['AMT_DRAWINGS_CURRENT'] / cc_bal['CNT_DRAWINGS_CURRENT']
cc_bal['RECEIVE_RATIO'] = cc_bal['AMT_RECIVABLE'] / cc_bal['AMT_RECEIVABLE_PRINCIPAL']
cc_bal['RECEIVE_PER'] = cc_bal['AMT_RECIVABLE'] / cc_bal['AMT_... | Home Credit Default Risk |
22,248,594 | df = train.drop(columns=['ACTION'])
train_x = train.drop(columns=['ACTION'])
train_y = train['ACTION']
test_x = test.drop(columns=['id'])
<split> | install = pd.read_csv('.. /input/home-credit-default-risk/installments_payments.csv')
install['PAY_PERCENT'] = install['AMT_INSTALMENT'] / install['AMT_PAYMENT']
install['PAY_DIFF'] = install['AMT_INSTALMENT'] - install['AMT_PAYMENT']
install['DPD'] = install['DAYS_ENTRY_PAYMENT'] - install['DAYS_INSTALMENT']
install[... | Home Credit Default Risk |
22,248,594 | X_train, X_test, y_train, y_test = train_test_split(train_x, train_y )<choose_model_class> | pos = pd.read_csv('.. /input/home-credit-default-risk/POS_CASH_balance.csv')
pos.columns = ['PC_'+ column if column !='SK_ID_CURR'
else column for column in pos.columns]
pos_num = pos.groupby(by = ['SK_ID_CURR'] ).agg(['max', 'mean', 'sum'] ).astype('float32')
test = test.merge(pos_num, on = ['SK_ID_CURR'], how = 'le... | Home Credit Default Risk |
22,248,594 | model = lm.LogisticRegression()
model.fit(X_train, y_train)
predictions = model.predict_proba(test_x)
<prepare_x_and_y> | prev = pd.read_csv('.. /input/home-credit-default-risk/previous_application.csv')
prev['DAYS_FIRST_DRAWING'].replace(365243, np.nan, inplace = True)
prev['DAYS_FIRST_DUE'].replace(365243, np.nan, inplace = True)
prev['DAYS_LAST_DUE_1ST_VERSION'].replace(365243, np.nan, inplace = True)
prev['DAYS_LAST_DUE'].replace(... | Home Credit Default Risk |
22,248,594 | print(f"{X_train.shape}, {X_test.shape}, {test.drop(columns=['id'] ).shape}")
test_x = test.drop(columns=['id'])
X_train.reset_index(drop=True, inplace=True)
y_train.reset_index(drop=True, inplace=True )<predict_on_test> | test['DAYS_EMPLOYED'].replace(365243, np.nan, inplace = True)
test['AGE'] = test['DAYS_BIRTH'] / - 365
test['AVG_EXT'] = test.iloc[:, 41:44].sum(axis=1)/(3- test.iloc[:,41:44].isnull().sum(axis=1))
test.EXT_SOURCE_1.fillna(test.AVG_EXT, inplace=True)
test.EXT_SOURCE_2.fillna(test.AVG_EXT, inplace=True)
test.EXT_SOUR... | Home Credit Default Risk |
22,248,594 | model = naive_bayes.CategoricalNB()
<train_model> | test['EmpAge_RATIO'] = test['DAYS_EMPLOYED'] / test['AGE']
test['CredInc_RATIO'] = test['AMT_CREDIT'] / test['AMT_INCOME_TOTAL']
test['AnnInc_RATIO'] = test['AMT_ANNUITY'] / test['AMT_INCOME_TOTAL']
test['AnnCred_RATIO'] = test['AMT_ANNUITY'] /(test['AMT_CREDIT'] + 1)
test['CredGoods_RATIO'] = test['AMT_CREDIT'] /(tes... | Home Credit Default Risk |
22,248,594 | model = DecisionTreeClassifier(max_depth=30)
clf = model.fit(X_train, y_train)
print(f'{clf.score(X_test,y_test)}')
predictions = clf.predict(test_x)
<train_model> | dels = ['APARTMENTS_MODE', 'BASEMENTAREA_MODE', 'YEARS_BEGINEXPLUATATION_MODE',
'YEARS_BUILD_MODE', 'COMMONAREA_MODE', 'ELEVATORS_MODE', 'ENTRANCES_MODE',
'FLOORSMAX_MODE', 'FLOORSMIN_MODE', 'LANDAREA_MODE', 'LIVINGAPARTMENTS_MODE',
'LIVINGAREA_MODE', 'NONLIVINGAPARTMENTS_MODE', 'NONLIVINGAREA_MODE',
'APARTMENTS_MEDI',... | Home Credit Default Risk |
22,248,594 | model = RandomForestClassifier(n_estimators = 300)
clf = model.fit(X_train, y_train)
print(f'{clf.score(X_test, y_test)}')
predictions = clf.predict_proba(test_x)
<define_variables> | test = test.replace([np.inf, -np.inf], np.nan ) | Home Credit Default Risk |
22,248,594 | feature_names = dict()
for column, name in enumerate(train):
if column == 0:
continue
feature_names[column - 1] = name
dataset_dir = './amazon'
create_cd(
label=0,
cat_features=list(range(1, train.columns.shape[0])) ,
feature_names=feature_names,
output_path=os.path.join(dataset_dir, 'train.cd')
)<prepare_x_and_y> | test_pred = LGBM_model.predict_proba(X_test)
print(test_pred.shape)
print(test_pred[:5] ) | Home Credit Default Risk |
22,248,594 | X = train.drop(columns=['ACTION'])
y = train.ACTION
cat_features = list(range(0, X.shape[1]))
print(cat_features )<define_variables> | submission = pd.read_csv('.. /input/home-credit-default-risk/sample_submission.csv')
submission.head(10 ) | Home Credit Default Risk |
22,248,594 | pool1 = Pool(data=X, label=y, cat_features=cat_features)
pool2 = Pool(
data=os.path.join('/kaggle/input/amazon-employee-access-challenge/', 'train.csv'),
delimiter=',',
column_description=os.path.join(dataset_dir, 'train.cd'),
has_header=True
)
pool3 = Pool(data=X, cat_features=cat_features)
X_prepared = X.values.... | submission.TARGET = test_pred[:,1]
submission.head(10 ) | Home Credit Default Risk |
22,248,594 | <train_model><EOS> | submission.to_csv('default_submission_wk06.csv', index=False, header = True ) | Home Credit Default Risk |
22,046,560 | <SOS> metric: AUC Kaggle data source: home-credit-default-risk<feature_engineering> | MainDir = ".. /input/.. /input/home-credit-default-risk"
test = pd.read_csv(f'{MainDir}/application_test.csv' ) | Home Credit Default Risk |
22,046,560 | model.get_feature_importance(prettified=True )<init_hyperparams> | preprocessor = joblib.load('.. /input/defaultdata08/default_preprocessor_08.joblib')
model = joblib.load('.. /input/defaultdata08/default_model_08.joblib')
print(type(model)) | Home Credit Default Risk |
22,046,560 | params = {}
params['loss_function'] = 'Logloss'
params['iterations'] = 93
params['custom_loss'] = 'AUC'
params['random_seed'] = 63
params['learning_rate'] = 0.5
cv_data = cv(
params = params,
pool = Pool(X, label=y, cat_features=cat_features),
fold_count=5,
shuffle=True,
partition_random_seed=0,
stratified=False,
verb... | bureau = pd.read_csv(f'{MainDir}/bureau.csv')
print(bureau.shape, "- shape of bureau table")
bureau_balance = pd.read_csv(f'{MainDir}/bureau_balance.csv')
bb_status = pd.crosstab(bureau_balance.SK_ID_BUREAU, bureau_balance.STATUS)
bb_status.columns = ['BB_'+column for column in bb_status.columns]
bureau = bureau.me... | Home Credit Default Risk |
22,046,560 | best_value = np.min(cv_data['test-Logloss-mean'])
best_iter = np.argmin(cv_data['test-Logloss-mean'])
print('Best validation Logloss score, not stratified: {:.4f}±{:.4f} on step {}'.format(
best_value,
cv_data['test-Logloss-std'][best_iter],
best_iter)
)<load_from_csv> | test_pred = model.predict_proba(X_test)
print(test_pred.shape)
print(test_pred[:5] ) | Home Credit Default Risk |
22,046,560 | <prepare_output><EOS> | submission = pd.read_csv('.. /input/home-credit-default-risk/sample_submission.csv')
submission.head(10)
submission.TARGET = test_pred[:,1]
submission.head(10)
submission.to_csv('default_submission_08.csv', index=False, header = True ) | Home Credit Default Risk |
19,576,721 | <SOS> metric: AUC Kaggle data source: home-credit-default-risk<save_to_csv> | import os
import gc
import numpy as np
import pandas as pd
from scipy.stats import kurtosis
from sklearn.metrics import roc_auc_score
from sklearn.preprocessing import MinMaxScaler
from sklearn.impute import SimpleImputer
from sklearn.linear_model import LogisticRegression
import matplotlib.pyplot as plt
import seaborn... | Home Credit Default Risk |
19,576,721 | os.chdir('/kaggle/working/')
os.curdir
sol = pd.DataFrame(predictions)
sol = sol.rename(columns={0:'Action'})
sol.index = range(1, 58922,1)
sol = sol.rename_axis('Id')
sol.to_csv('submission.csv' )<load_pretrained> | DATA_DIRECTORY = ".. /input/home-credit-loan-better-data-processing" | Home Credit Default Risk |
19,576,721 | with zipfile.ZipFile('/kaggle/input/sentiment-analysis-on-movie-reviews/train.tsv.zip','r')as zip_ref:
zip_ref.extractall("./sentiment-analysis-on-movie-reviews/")
with zipfile.ZipFile('/kaggle/input/sentiment-analysis-on-movie-reviews/test.tsv.zip','r')as zip_ref:
zip_ref.extractall("./sentiment-analysis-on-movie-rev... | train = pd.read_csv(os.path.join(DATA_DIRECTORY, 'train.csv'))
test = pd.read_csv(os.path.join(DATA_DIRECTORY, 'test.csv'))
labels = pd.read_csv(os.path.join(DATA_DIRECTORY, 'labels.csv'))
| Home Credit Default Risk |
19,576,721 | data_source=pd.read_table("/kaggle/working/sentiment-analysis-on-movie-reviews/train.tsv",sep='\t')
data_source=data_source[['Phrase','Sentiment']].copy()
data_source<string_transform> | train = train.rename(columns = lambda x:re.sub('[^A-Za-z0-9_]+', '', x))
test = test.rename(columns = lambda x:re.sub('[^A-Za-z0-9_]+', '', x))
labels = labels.rename(columns = lambda x:re.sub('[^A-Za-z0-9_]+', '', x)) | Home Credit Default Risk |
19,576,721 | dff=[len(i.split(" ")) for i in data_source.Phrase[:10]]
max(dff )<import_modules> | train=np.nan_to_num(train)
test=np.nan_to_num(test)
labels=np.nan_to_num(labels ) | Home Credit Default Risk |
19,576,721 | from transformers import TFBertModel, BertConfig, BertTokenizerFast, TFAutoModel
from tensorflow.keras.layers import Input, Dropout, Dense
from tensorflow.keras.models import Model
from tensorflow.keras.optimizers import Adam
from tensorflow.keras.callbacks import EarlyStopping
from tensorflow.keras.initializers import... | train = pd.DataFrame(train)
test = pd.DataFrame(test)
labels=pd.DataFrame(labels ) | Home Credit Default Risk |
19,576,721 | X_train_data, X_validation_data, y_train_data, y_validation_data = train_test_split(data_source.index.values,
data_source.Sentiment.values,
test_size=0.10,
random_state=42,
stratify=data_source.Sentiment)
<feature_engineering> | X_train, X_test, y_train, y_test = train_test_split(train, labels, random_state=42 ) | Home Credit Default Risk |
19,576,721 | data_source['data_type'] = ['not_set']*data_source.shape[0]
data_source.loc[X_train_data, 'data_type'] = 'training'
data_source.loc[X_validation_data, 'data_type'] = 'validation'<count_missing_values> | clf = DummyClassifier(strategy= 'most_frequent' ).fit(X_train,y_train)
y_pred = clf.predict(X_test)
print('y actual :
' + str(y_test.value_counts()))
print('y predicted :
' + str(pd.Series(y_pred ).value_counts())) | Home Credit Default Risk |
19,576,721 | data_source.isnull().sum()<filter> | print('Accuracy Score : ' + str(accuracy_score(y_test,y_pred)))
print('Precision Score : ' + str(precision_score(y_test,y_pred)))
print('Recall Score : ' + str(recall_score(y_test,y_pred)))
print('F1 Score : ' + str(f1_score(y_test,y_pred)))
print('Confusion Matrix :
' + str(confusion_matrix(y_test,y_pred)) ) | Home Credit Default Risk |
19,576,721 | data_source[data_source.data_type=='training'].Phrase<load_pretrained> | clf = LGBMClassifier().fit(X_train,y_train)
y_pred = clf.predict(X_test)
print('Accuracy Score : ' + str(accuracy_score(y_test,y_pred)))
print('Precision Score : ' + str(precision_score(y_test,y_pred)))
print('Recall Score : ' + str(recall_score(y_test,y_pred)))
print('F1 Score : ' + str(f1_score(y_test,y_pred)))
... | Home Credit Default Risk |
19,576,721 | max_token_length = max(dff)+3
number_of_samples = len(data_source)
bert = 'bert-base-cased'
config = BertConfig.from_pretrained(bert)
config.output_hidden_states = False
tokenizer = BertTokenizerFast.from_pretrained(pretrained_model_name_or_path = bert, config = config)
<categorify> | clf = LogisticRegression()
grid_values = {'penalty': ['l2'],'C':[0.001,.009,0.01,.09,1,5,10,25]}
grid_clf_acc = GridSearchCV(clf, param_grid = grid_values,scoring = 'recall')
grid_clf_acc.fit(X_train, y_train)
y_pred_acc = grid_clf_acc.predict(X_test)
print('Accuracy Score : ' + str(accuracy_score(y_test,y_pred_acc)... | Home Credit Default Risk |
19,576,721 | def map_function(input_ids, masks,labels):
return {'input_ids': input_ids, 'attention_mask': masks},labels<categorify> | pred = model.predict_proba(df_test ) | Home Credit Default Risk |
19,576,721 | <categorify><EOS> | submit = test[['SK_ID_CURR']]
submit['TARGET'] = pred
submit.to_csv('lgbm_Minimized_code.csv', index = False ) | Home Credit Default Risk |
18,348,927 | <SOS> metric: AUC Kaggle data source: home-credit-default-risk<define_variables> | warnings.simplefilter(action='ignore', category=FutureWarning ) | Home Credit Default Risk |
18,348,927 | batch_size = 32
train_dataset = train_dataset.shuffle(1000 ).batch(batch_size, drop_remainder=True )<categorify> | DATA_DIRECTORY = ".. /input/home-credit-default-risk" | Home Credit Default Risk |
18,348,927 | y_senti = to_categorical(data_source[data_source.data_type=='validation'].Sentiment)
v= tokenizer(
text=data_source[data_source.data_type=='validation'].Phrase.to_list() ,
add_special_tokens=True,
max_length=max_token_length,
truncation=True,
padding=True,
return_tensors='tf',
return_token_type_ids = False,
return_at... | df_train = pd.read_csv(os.path.join(DATA_DIRECTORY, 'application_train.csv'))
df_test = pd.read_csv(os.path.join(DATA_DIRECTORY, 'application_test.csv'))
df = df_train.append(df_test)
del df_train, df_test; gc.collect() | Home Credit Default Risk |
18,348,927 | input_ids = tf.keras.Input(shape=(max_token_length,), name='input_ids', dtype='int32')
attention_mask = tf.keras.Input(shape=(max_token_length,), name='attention_mask', dtype='int32')
inputs = {'input_ids': input_ids, 'attention_mask': attention_mask}
bert = TFAutoModel.from_pretrained('bert-base-cased')
embeddings ... | df = df[df['AMT_INCOME_TOTAL'] < 20000000]
df = df[df['CODE_GENDER'] != 'XNA']
df['DAYS_EMPLOYED'].replace(365243, np.nan, inplace=True)
df['DAYS_LAST_PHONE_CHANGE'].replace(0, np.nan, inplace=True ) | Home Credit Default Risk |
18,348,927 | optimizer = tf.keras.optimizers.Adam(lr=1e-5, decay=1e-6)
loss = tf.keras.losses.CategoricalCrossentropy()
accuracy = tf.keras.metrics.CategoricalAccuracy('accuracy' )<choose_model_class> | def get_age_group(days_birth):
age_years = -days_birth / 365
if age_years < 27: return 1
elif age_years < 40: return 2
elif age_years < 50: return 3
elif age_years < 65: return 4
elif age_years < 99: return 5
else: return 0 | Home Credit Default Risk |
18,348,927 | model.compile(optimizer=optimizer, loss=loss, metrics=[accuracy] )<train_model> | docs = [f for f in df.columns if 'FLAG_DOC' in f]
df['DOCUMENT_COUNT'] = df[docs].sum(axis=1)
df['NEW_DOC_KURT'] = df[docs].kurtosis(axis=1)
df['AGE_RANGE'] = df['DAYS_BIRTH'].apply(lambda x: get_age_group(x)) | Home Credit Default Risk |
18,348,927 | history1 = model.fit(
train_dataset,
validation_data= validation_dataset,
epochs=8 )<save_model> | df['EXT_SOURCES_PROD'] = df['EXT_SOURCE_1'] * df['EXT_SOURCE_2'] * df['EXT_SOURCE_3']
df['EXT_SOURCES_WEIGHTED'] = df.EXT_SOURCE_1 * 2 + df.EXT_SOURCE_2 * 1 + df.EXT_SOURCE_3 * 3
np.warnings.filterwarnings('ignore', r'All-NaN(slice|axis)encountered')
for function_name in ['min', 'max', 'mean', 'nanmedian', 'var']:
fea... | Home Credit Default Risk |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.