def main(params: dict):
import mlflow
print("start params={}".format(params))
df = pd.read_pickle(
"../input/riiid-test-answer-prediction/train_merged.pickle")
print(df.head(10))
# df = pd.read_pickle("../input/riiid-test-answer-prediction/split10/train_0.pickle").sort_values(["user_id", "timestamp"]).reset_index(drop=True)
if is_debug:
df = df.head(30000)
logger = get_logger()
feature_factory_dict = {"user_id": {}}
feature_factory_dict["user_id"][
"DurationPreviousContent"] = DurationPreviousContent(
is_partial_fit=True)
feature_factory_manager = FeatureFactoryManager(
feature_factory_dict=feature_factory_dict,
logger=logger,
split_num=1,
model_id="all",
load_feature=not is_debug,
save_feature=not is_debug)
print("all_predict")
df = feature_factory_manager.all_predict(df)
df = df[[
"user_id", "content_id", "content_type_id", "part", "user_answer",
"answered_correctly", "prior_question_elapsed_time",
"duration_previous_content"
]]
print("data preprocess")
df["prior_question_elapsed_time_bin300"] = [
x // 1000 if x // 1000 < 300 else 300
for x in df["prior_question_elapsed_time"].fillna(0).values
]
df["duration_previous_content_bin300"] = [
x // 1000 if x // 1000 < 300 else 300
for x in df["duration_previous_content"].fillna(0).values
]
train_idx = []
val_idx = []
np.random.seed(0)
for _, w_df in df[df["content_type_id"] == 0].groupby("user_id"):
if np.random.random() < 0.1:
# all val
val_idx.extend(w_df.index.tolist())
else:
train_num = int(len(w_df) * 0.9)
train_idx.extend(w_df[:train_num].index.tolist())
val_idx.extend(w_df[train_num:].index.tolist())
ff_for_transformer = FeatureFactoryForTransformer(
column_config={
("content_id", "content_type_id"): {
"type": "category"
},
"user_answer": {
"type": "category"
},
"part": {
"type": "category"
},
"prior_question_elapsed_time_bin300": {
"type": "category"
},
"duration_previous_content_bin300": {
"type": "category"
}
},
dict_path="../feature_engineering/",
sequence_length=params["max_seq"],
logger=logger)
ff_for_transformer.make_dict(df)
n_skill = len(ff_for_transformer.embbed_dict[("content_id",
"content_type_id")])
if not load_pickle:
df["is_val"] = 0
df["is_val"].loc[val_idx] = 1
w_df = df[df["is_val"] == 0]
w_df["group"] = (
w_df.groupby("user_id")["user_id"].transform("count") -
w_df.groupby("user_id").cumcount()) // params["max_seq"]
w_df["user_id"] = w_df["user_id"].astype(
str) + "_" + w_df["group"].astype(str)
group = ff_for_transformer.all_predict(w_df)
dataset_train = SAKTDataset(group,
n_skill=n_skill,
max_seq=params["max_seq"])
del w_df
gc.collect()
ff_for_transformer = FeatureFactoryForTransformer(
column_config={
("content_id", "content_type_id"): {
"type": "category"
},
"user_answer": {
"type": "category"
},
"part": {
"type": "category"
},
"prior_question_elapsed_time_bin300": {
"type": "category"
},
"duration_previous_content_bin300": {
"type": "category"
}
},
dict_path="../feature_engineering/",
sequence_length=params["max_seq"],
logger=logger)
if not load_pickle:
group = ff_for_transformer.all_predict(df[df["content_type_id"] == 0])
dataset_val = SAKTDataset(group,
is_test=True,
n_skill=n_skill,
max_seq=params["max_seq"])
os.makedirs("../input/feature_engineering/model048", exist_ok=True)
if not is_debug and not load_pickle:
with open(f"../input/feature_engineering/model048/train.pickle",
"wb") as f:
pickle.dump(dataset_train, f)
with open(f"../input/feature_engineering/model048/val.pickle",
"wb") as f:
pickle.dump(dataset_val, f)
if load_pickle:
with open(f"../input/feature_engineering/model048/train.pickle",
"rb") as f:
dataset_train = pickle.load(f)
with open(f"../input/feature_engineering/model048/val.pickle",
"rb") as f:
dataset_val = pickle.load(f)
dataloader_train = DataLoader(dataset_train,
batch_size=512,
shuffle=True,
num_workers=1)
dataloader_val = DataLoader(dataset_val,
batch_size=512,
shuffle=False,
num_workers=1)
model = SAKTModel(n_skill,
embed_dim=params["embed_dim"],
max_seq=params["max_seq"],
dropout=dropout)
optimizer = torch.optim.Adam(model.parameters(), lr=params["lr"])
criterion = nn.BCEWithLogitsLoss()
model.to(device)
criterion.to(device)
for epoch in range(epochs):
loss, acc, auc, auc_val = train_epoch(model, dataloader_train,
dataloader_val, optimizer,
criterion, device)
print("epoch - {} train_loss - {:.3f} auc - {:.4f} auc-val: {:.4f}".
format(epoch, loss, auc, auc_val))
preds = []
labels = []
for item in tqdm(dataloader_val):
x = item["x"].to(device).long()
target_id = item["target_id"].to(device).long()
part = item["part"].to(device).long()
label = item["label"].to(device).float()
elapsed_time = item["elapsed_time"].to(device).long()
duration_previous_content = item["duration_previous_content"].to(
device).long()
output, atten_weight = model(x, target_id, part, elapsed_time,
duration_previous_content)
preds.extend(torch.nn.Sigmoid()(
output[:, -1]).view(-1).data.cpu().numpy().tolist())
labels.extend(label[:, -1].view(-1).data.cpu().numpy().tolist())
df_oof = pd.DataFrame()
df_oof["row_id"] = df.loc[val_idx].index
df_oof["predict"] = preds
df_oof["target"] = df.loc[val_idx]["answered_correctly"].values
auc_transformer = roc_auc_score(df_oof["target"].values,
df_oof["predict"].values)
print("single transformer: {:.4f}".format(auc_transformer))
df_oof.to_csv(f"{output_dir}/transformers1.csv", index=False)
df_oof2 = pd.read_csv(
"../output/ex_237/20201213110353/oof_train_0_lgbm.csv")
df_oof2.columns = ["row_id", "predict_lgbm", "target"]
df_oof2 = pd.merge(df_oof, df_oof2, how="inner")
auc_lgbm = roc_auc_score(df_oof2["target"].values,
df_oof2["predict_lgbm"].values)
print("lgbm: {:.4f}".format(auc_lgbm))
print("ensemble")
max_auc = 0
max_nn_ratio = 0
for r in np.arange(0, 1.05, 0.05):
auc = roc_auc_score(
df_oof2["target"].values, df_oof2["predict_lgbm"].values *
(1 - r) + df_oof2["predict"].values * r)
print("[nn_ratio: {:.2f}] AUC: {:.4f}".format(r, auc))
if max_auc < auc:
max_auc = auc
max_nn_ratio = r
print(len(df_oof2))
if not is_debug:
mlflow.start_run(experiment_id=10, run_name=os.path.basename(__file__))
mlflow.log_param("count_row", len(df))
for key, value in params.items():
mlflow.log_param(key, value)
mlflow.log_metric("auc_val", auc_transformer)
mlflow.log_metric("auc_lgbm", auc_lgbm)
mlflow.log_metric("auc_ensemble", max_auc)
mlflow.log_metric("ensemble_nn_ratio", max_nn_ratio)
mlflow.end_run()
torch.save(model.state_dict(), f"{output_dir}/transformers.pth")
with open(f"{output_dir}/transformer_param.json", "w") as f:
json.dump(params, f)
if is_make_feature_factory:
ff_for_transformer = FeatureFactoryForTransformer(
column_config={
("content_id", "content_type_id"): {
"type": "category"
},
"part": {
"type": "category"
}
},
dict_path="../feature_engineering/",
sequence_length=params["max_seq"],
logger=logger)
df = pd.read_pickle(
"../input/riiid-test-answer-prediction/train_merged.pickle")
if is_debug:
df = df.head(10000)
df = df.sort_values(["user_id", "timestamp"]).reset_index(drop=True)
ff_for_transformer.fit(df)
ff_for_transformer.logger = None
with open(
f"{output_dir}/feature_factory_manager_for_transformer.pickle",
"wb") as f:
pickle.dump(ff_for_transformer, f)
def main(params: dict, output_dir: str):
import mlflow
print("start params={}".format(params))
model_id = "train_0"
logger = get_logger()
# df = pd.read_pickle("../input/riiid-test-answer-prediction/train_merged.pickle")
df = pd.read_pickle(
"../input/riiid-test-answer-prediction/split10/train_0.pickle"
).sort_values(["user_id", "timestamp"]).reset_index(drop=True)
if is_debug:
df = df.head(30000)
df["prior_question_had_explanation"] = df[
"prior_question_had_explanation"].fillna(-1)
df["answered_correctly"] = df["answered_correctly"].replace(-1, np.nan)
column_config = {
("content_id", "content_type_id"): {
"type": "category"
},
"user_answer": {
"type": "leakage_feature"
},
"answered_correctly": {
"type": "leakage_feature"
},
"part": {
"type": "category"
},
"prior_question_elapsed_time_bin300": {
"type": "category"
},
"duration_previous_content_bin300": {
"type": "category"
},
"prior_question_had_explanation": {
"type": "category"
},
"rating_diff_content_user_id": {
"type": "numeric"
},
"task_container_id_bin300": {
"type": "category"
},
"previous_answer_index_question_id": {
"type": "category"
},
"previous_answer_question_id": {
"type": "category"
},
"timediff-elapsedtime_bin500": {
"type": "category"
},
"timedelta_log10": {
"type": "category"
}
}
if not load_pickle or is_debug:
feature_factory_dict = {"user_id": {}}
feature_factory_dict["user_id"][
"DurationPreviousContent"] = DurationPreviousContent(
is_partial_fit=True)
feature_factory_dict["user_id"][
"ElapsedTimeBinningEncoder"] = ElapsedTimeBinningEncoder()
feature_factory_dict["user_id"][
"UserContentRateEncoder"] = UserContentRateEncoder(
rate_func="elo", column="user_id")
feature_factory_dict["user_id"]["PreviousAnswer2"] = PreviousAnswer2(
groupby="user_id",
column="question_id",
is_debug=is_debug,
model_id=model_id,
n=300)
feature_factory_dict["user_id"][
"StudyTermEncoder2"] = StudyTermEncoder2(is_partial_fit=True)
feature_factory_dict["user_id"][
f"MeanAggregatorStudyTimebyUserId"] = MeanAggregator(
column="user_id", agg_column="study_time", remove_now=False)
feature_factory_dict["user_id"][
"ElapsedTimeMeanByContentIdEncoder"] = ElapsedTimeMeanByContentIdEncoder(
)
feature_factory_dict["post"] = {
"DurationFeaturePostProcess": DurationFeaturePostProcess()
}
feature_factory_manager = FeatureFactoryManager(
feature_factory_dict=feature_factory_dict,
logger=logger,
split_num=1,
model_id=model_id,
load_feature=not is_debug,
save_feature=not is_debug)
print("all_predict")
df = feature_factory_manager.all_predict(df)
def f(x):
x = x // 1000
if x < -100:
return -100
if x > 400:
return 400
return x
df["task_container_id_bin300"] = [
x if x < 300 else 300 for x in df["task_container_id"]
]
df["timediff-elapsedtime_bin500"] = [
f(x) for x in df["timediff-elapsedtime"].values
]
df["timedelta_log10"] = np.log10(
df["duration_previous_content"].values)
df["timedelta_log10"] = df["timedelta_log10"].replace(
-np.inf, -1).replace(np.inf, -1).fillna(-1).astype("int8")
df = df[[
"user_id", "content_id", "content_type_id", "part", "user_answer",
"answered_correctly", "prior_question_elapsed_time_bin300",
"duration_previous_content_bin300",
"prior_question_had_explanation", "rating_diff_content_user_id",
"task_container_id_bin300", "previous_answer_index_question_id",
"previous_answer_question_id", "row_id",
"timediff-elapsedtime_bin500", "timedelta_log10"
]]
print(df.head(10))
print("data preprocess")
ff_for_transformer = FeatureFactoryForTransformer(
column_config=column_config,
dict_path="../feature_engineering/",
sequence_length=params["max_seq"],
logger=logger)
ff_for_transformer.make_dict(df=df)
n_skill = len(ff_for_transformer.embbed_dict[("content_id",
"content_type_id")])
if not load_pickle or is_debug:
df_val_row = pd.read_feather(
"../../riiid_takoi/notebook/fe/validation_row_id.feather").head(
len(df))
if is_debug:
df_val_row = df_val_row.head(3000)
df_val_row["is_val"] = 1
df = pd.merge(df, df_val_row, how="left", on="row_id")
df["is_val"] = df["is_val"].fillna(0)
print(df["is_val"].value_counts())
w_df = df[df["is_val"] == 0]
w_df["group"] = (
w_df.groupby("user_id")["user_id"].transform("count") -
w_df.groupby("user_id").cumcount()) // params["max_seq"]
w_df["user_id"] = w_df["user_id"].astype(
str) + "_" + w_df["group"].astype(str)
group = ff_for_transformer.all_predict(w_df)
dataset_train = SAKTDataset(group,
n_skill=n_skill,
max_seq=params["max_seq"])
del w_df
gc.collect()
ff_for_transformer = FeatureFactoryForTransformer(
column_config=column_config,
dict_path="../feature_engineering/",
sequence_length=params["max_seq"],
logger=logger)
if not load_pickle or is_debug:
group = ff_for_transformer.all_predict(df[df["content_type_id"] == 0])
dataset_val = SAKTDataset(group,
is_test=True,
n_skill=n_skill,
max_seq=params["max_seq"])
os.makedirs("../input/feature_engineering/model270", exist_ok=True)
if not is_debug and not load_pickle:
with open(f"../input/feature_engineering/model270/train.pickle",
"wb") as f:
pickle.dump(dataset_train, f)
with open(f"../input/feature_engineering/model270/val.pickle",
"wb") as f:
pickle.dump(dataset_val, f)
if not is_debug and load_pickle:
with open(f"../input/feature_engineering/model270/train.pickle",
"rb") as f:
dataset_train = pickle.load(f)
with open(f"../input/feature_engineering/model270/val.pickle",
"rb") as f:
dataset_val = pickle.load(f)
print("loaded!")
dataloader_train = DataLoader(dataset_train,
batch_size=params["batch_size"],
shuffle=True)
dataloader_val = DataLoader(dataset_val,
batch_size=params["batch_size"],
shuffle=False)
model = SAKTModel(n_skill,
embed_dim=params["embed_dim"],
max_seq=params["max_seq"],
dropout=dropout,
cont_emb=params["cont_emb"])
param_optimizer = list(model.named_parameters())
no_decay = ['bias', 'LayerNorm.bias', 'LayerNorm.weight']
optimizer_grouped_parameters = [{
'params':
[p for n, p in param_optimizer if not any(nd in n for nd in no_decay)],
'weight_decay':
0.0
}, {
'params':
[p for n, p in param_optimizer if any(nd in n for nd in no_decay)],
'weight_decay':
0.0
}]
optimizer = AdamW(
optimizer_grouped_parameters,
lr=params["lr"],
weight_decay=0.0,
)
num_train_optimization_steps = int(len(dataloader_train) * 20)
scheduler = get_linear_schedule_with_warmup(
optimizer,
num_warmup_steps=params["num_warmup_steps"],
num_training_steps=num_train_optimization_steps)
criterion = nn.BCEWithLogitsLoss()
model.to(device)
criterion.to(device)
for epoch in range(epochs):
loss, acc, auc, auc_val = train_epoch(model, dataloader_train,
dataloader_val, optimizer,
criterion, scheduler, epoch,
device)
print("epoch - {} train_loss - {:.3f} auc - {:.4f} auc-val: {:.4f}".
format(epoch, loss, auc, auc_val))
preds = []
labels = []
with torch.no_grad():
for item in tqdm(dataloader_val):
label = item["label"].to(device).float()
output = model(item, device)
preds.extend(torch.nn.Sigmoid()(
output[:, -1]).view(-1).data.cpu().numpy().tolist())
labels.extend(label[:, -1].view(-1).data.cpu().numpy().tolist())
auc_transformer = roc_auc_score(labels, preds)
print("single transformer: {:.4f}".format(auc_transformer))
df_oof = pd.DataFrame()
# df_oof["row_id"] = df.loc[val_idx].index
print(len(dataloader_val))
print(len(preds))
df_oof["predict"] = preds
df_oof["target"] = labels
df_oof.to_csv(f"{output_dir}/transformers1.csv", index=False)
"""
df_oof2 = pd.read_csv("../output/ex_237/20201213110353/oof_train_0_lgbm.csv")
df_oof2.columns = ["row_id", "predict_lgbm", "target"]
df_oof2 = pd.merge(df_oof, df_oof2, how="inner")
auc_lgbm = roc_auc_score(df_oof2["target"].values, df_oof2["predict_lgbm"].values)
print("lgbm: {:.4f}".format(auc_lgbm))
print("ensemble")
max_auc = 0
max_nn_ratio = 0
for r in np.arange(0, 1.05, 0.05):
auc = roc_auc_score(df_oof2["target"].values, df_oof2["predict_lgbm"].values*(1-r) + df_oof2["predict"].values*r)
print("[nn_ratio: {:.2f}] AUC: {:.4f}".format(r, auc))
if max_auc < auc:
max_auc = auc
max_nn_ratio = r
print(len(df_oof2))
"""
if not is_debug:
mlflow.start_run(experiment_id=10, run_name=os.path.basename(__file__))
for key, value in params.items():
mlflow.log_param(key, value)
mlflow.log_metric("auc_val", auc_transformer)
mlflow.end_run()
torch.save(model.state_dict(), f"{output_dir}/transformers.pth")
del model
torch.cuda.empty_cache()
with open(f"{output_dir}/transformer_param.json", "w") as f:
json.dump(params, f)
if is_make_feature_factory:
# feature factory
feature_factory_dict = {"user_id": {}}
feature_factory_dict["user_id"][
"DurationPreviousContent"] = DurationPreviousContent(
is_partial_fit=True)
feature_factory_dict["user_id"][
"ElapsedTimeBinningEncoder"] = ElapsedTimeBinningEncoder()
feature_factory_manager = FeatureFactoryManager(
feature_factory_dict=feature_factory_dict,
logger=logger,
split_num=1,
model_id="all",
load_feature=not is_debug,
save_feature=not is_debug)
ff_for_transformer = FeatureFactoryForTransformer(
column_config=column_config,
dict_path="../feature_engineering/",
sequence_length=params["max_seq"],
logger=logger)
df = pd.read_pickle(
"../input/riiid-test-answer-prediction/train_merged.pickle")
if is_debug:
df = df.head(10000)
df = df.sort_values(["user_id", "timestamp"]).reset_index(drop=True)
feature_factory_manager.fit(df)
df = feature_factory_manager.all_predict(df)
for dicts in feature_factory_manager.feature_factory_dict.values():
for factory in dicts.values():
factory.logger = None
feature_factory_manager.logger = None
with open(f"{output_dir}/feature_factory_manager.pickle", "wb") as f:
pickle.dump(feature_factory_manager, f)
ff_for_transformer.fit(df)
ff_for_transformer.logger = None
with open(
f"{output_dir}/feature_factory_manager_for_transformer.pickle",
"wb") as f:
pickle.dump(ff_for_transformer, f)
def make_feature_factory_manager(split_num):
logger = get_logger()
feature_factory_dict = {}
feature_factory_dict["tags"] = {"TagsSeparator": TagsSeparator()}
for column in [
"content_id", "user_id", "part", "prior_question_had_explanation",
"tags1", "tags2", ("user_id", "prior_question_had_explanation"),
("user_id", "part"), ("content_id", "prior_question_had_explanation")
]:
is_partial_fit = (column == "content_id" or column == "user_id")
if type(column) == str:
feature_factory_dict[column] = {
"CountEncoder":
CountEncoder(column=column, is_partial_fit=is_partial_fit),
"TargetEncoder":
TargetEncoder(column=column, is_partial_fit=is_partial_fit)
}
else:
feature_factory_dict[column] = {
"CountEncoder":
CountEncoder(column=list(column),
is_partial_fit=is_partial_fit),
"TargetEncoder":
TargetEncoder(column=list(column),
is_partial_fit=is_partial_fit)
}
feature_factory_dict["user_id"][
"ShiftDiffEncoderTimestamp"] = ShiftDiffEncoder(groupby="user_id",
column="timestamp",
is_partial_fit=True)
feature_factory_dict["user_id"][
"ShiftDiffEncoderContentId"] = ShiftDiffEncoder(groupby="user_id",
column="content_id")
for column in ["user_id", "content_id"]:
feature_factory_dict[column][
f"MeanAggregatorPriorQuestionElapsedTimeby{column}"] = MeanAggregator(
column=column,
agg_column="prior_question_elapsed_time",
remove_now=True)
feature_factory_dict["user_id"][
"UserLevelEncoder2ContentId"] = UserLevelEncoder2(
vs_column="content_id")
feature_factory_dict["user_id"][
"UserCountBinningEncoder"] = UserCountBinningEncoder(
is_partial_fit=True)
feature_factory_dict["user_count_bin"] = {}
feature_factory_dict["user_count_bin"]["CountEncoder"] = CountEncoder(
column="user_count_bin")
feature_factory_dict["user_count_bin"]["TargetEncoder"] = TargetEncoder(
column="user_count_bin")
feature_factory_dict[("user_id", "user_count_bin")] = {
"CountEncoder": CountEncoder(column=["user_id", "user_count_bin"]),
"TargetEncoder": TargetEncoder(column=["user_id", "user_count_bin"])
}
feature_factory_dict[("content_id", "user_count_bin")] = {
"CountEncoder": CountEncoder(column=["content_id", "user_count_bin"]),
"TargetEncoder": TargetEncoder(column=["content_id", "user_count_bin"])
}
feature_factory_dict[(
"prior_question_had_explanation", "user_count_bin")] = {
"CountEncoder":
CountEncoder(
column=["prior_question_had_explanation", "user_count_bin"]),
"TargetEncoder":
TargetEncoder(
column=["prior_question_had_explanation", "user_count_bin"])
}
feature_factory_dict["user_id"][
"CategoryLevelEncoderPart"] = CategoryLevelEncoder(
groupby_column="user_id", agg_column="part", categories=[2, 5])
feature_factory_dict["user_count_bin"]["CategoryLevelEncoderUserCountBin"] = \
CategoryLevelEncoder(groupby_column="user_id",
agg_column="user_count_bin",
categories=[0])
feature_factory_dict["prior_question_elapsed_time"] = {
"PriorQuestionElapsedTimeBinningEncoder":
PriorQuestionElapsedTimeBinningEncoder()
}
feature_factory_dict[("part", "prior_question_elapsed_time_bin")] = {
"CountEncoder":
CountEncoder(column=["part", "prior_question_elapsed_time_bin"]),
"TargetEncoder":
TargetEncoder(column=["part", "prior_question_elapsed_time_bin"])
}
feature_factory_dict["user_id"]["PreviousAnswer2"] = PreviousAnswer2(
groupby="user_id", column="content_id")
feature_factory_manager = FeatureFactoryManager(
feature_factory_dict=feature_factory_dict,
logger=logger,
split_num=split_num)
return feature_factory_manager
def make_feature_factory_manager(split_num, model_id=None):
logger = get_logger()
feature_factory_dict = {}
for column in ["user_id", "content_id"]:
is_partial_fit = (column == "content_id" or column == "user_id")
if type(column) == str:
feature_factory_dict[column] = {
"TargetEncoder": TargetEncoder(column=column, is_partial_fit=is_partial_fit)
}
else:
feature_factory_dict[column] = {
"TargetEncoder": TargetEncoder(column=list(column), is_partial_fit=is_partial_fit)
}
feature_factory_dict["user_id"]["ShiftDiffEncoderTimestamp"] = ShiftDiffEncoder(groupby="user_id",
column="timestamp",
is_partial_fit=True)
feature_factory_dict["user_id"]["PastNTimestampEncoder"] = PastNFeatureEncoder(column="timestamp",
past_ns=[5],
agg_funcs=["vslast"],
remove_now=False)
feature_factory_dict["user_id"]["StudyTermEncoder"] = StudyTermEncoder(is_partial_fit=True)
feature_factory_dict["user_id"]["ElapsedTimeVsShiftDiffEncoder"] = ElapsedTimeVsShiftDiffEncoder()
feature_factory_dict["user_id"]["CountEncoder"] = CountEncoder(column="user_id", is_partial_fit=True)
feature_factory_dict["user_id"]["UserCountBinningEncoder"] = UserCountBinningEncoder(is_partial_fit=True)
feature_factory_dict["user_count_bin"] = {}
feature_factory_dict["user_count_bin"]["TargetEncoder"] = TargetEncoder(column="user_count_bin")
feature_factory_dict[("user_id", "user_count_bin")] = {
"TargetEncoder": TargetEncoder(column=["user_id", "user_count_bin"])
}
feature_factory_dict[("content_id", "user_count_bin")] = {
"TargetEncoder": TargetEncoder(column=["content_id", "user_count_bin"])
}
feature_factory_dict[("user_id", "part")] = {
"UserContentRateEncoder": UserContentRateEncoder(column=["user_id", "part"],
rate_func="elo")
}
for column in ["user_id", "content_id", "part", ("user_id", "part")]:
if column not in feature_factory_dict:
feature_factory_dict[column] = {}
if type(column) == str:
feature_factory_dict[column][f"MeanAggregatorShiftDiffTimeElapsedTimeby{column}"] = MeanAggregator(column=column,
agg_column="shiftdiff_timestamp_by_user_id_cap200k",
remove_now=True)
feature_factory_dict[column][f"MeanAggregatorStudyTimeby{column}"] = MeanAggregator(column=column,
agg_column="study_time",
remove_now=True)
else:
feature_factory_dict[column][f"MeanAggregatorShiftDiffTimeElapsedTimeby{column}"] = MeanAggregator(column=list(column),
agg_column="shiftdiff_timestamp_by_user_id_cap200k",
remove_now=True)
feature_factory_dict[column][f"MeanAggregatorStudyTimeby{column}"] = MeanAggregator(column=list(column),
agg_column="study_time",
remove_now=True)
feature_factory_dict["user_id"]["CategoryLevelEncoderPart"] = CategoryLevelEncoder(groupby_column="user_id",
agg_column="part",
categories=[2, 5])
feature_factory_dict["prior_question_elapsed_time"] = {
"PriorQuestionElapsedTimeBinningEncoder": PriorQuestionElapsedTimeBinningEncoder(is_partial_fit=True)
}
feature_factory_dict[("part", "prior_question_elapsed_time_bin")] = {
"TargetEncoder": TargetEncoder(column=["part", "prior_question_elapsed_time_bin"])
}
feature_factory_dict["user_id"]["PreviousAnswer2"] = PreviousAnswer2(groupby="user_id",
column="content_id",
is_debug=is_debug,
model_id=model_id,
n=300)
feature_factory_dict["user_id"]["PreviousNAnsweredCorrectly"] = PreviousNAnsweredCorrectly(n=3,
is_partial_fit=True)
feature_factory_dict["user_id"]["Counter"] = Counter(groupby_column="user_id",
agg_column="prior_question_had_explanation",
categories=[0, 1])
feature_factory_dict[f"previous_3_ans"] = {
"TargetEncoder": TargetEncoder(column="previous_3_ans")
}
feature_factory_dict["user_id"]["QuestionLectureTableEncoder2"] = QuestionLectureTableEncoder2(model_id=model_id,
is_debug=is_debug,
past_n=100,
min_size=100)
feature_factory_dict["user_id"]["QuestionQuestionTableEncoder2"] = QuestionQuestionTableEncoder2(model_id=model_id,
is_debug=is_debug,
past_n=100,
min_size=300)
feature_factory_dict["user_id"]["UserContentRateEncoder"] = UserContentRateEncoder(column="user_id",
rate_func="elo")
feature_factory_dict["post"] = {
"ContentIdTargetEncoderAggregator": TargetEncoderAggregator()
}
feature_factory_manager = FeatureFactoryManager(feature_factory_dict=feature_factory_dict,
logger=logger,
split_num=split_num,
model_id=model_id,
load_feature=not is_debug,
save_feature=not is_debug)
return feature_factory_manager
feature_factory_dict["content_id"][
"MeanAggregatorPriorQuestionElapsedTime"] = MeanAggregator(
column="content_id",
agg_column="prior_question_elapsed_time",
remove_now=True)
for column in [("user_id", "content_type_id"),
("user_id", "prior_question_had_explanation"),
("user_id", "tag"), ("user_id", "part"),
("content_type_id", "part")]:
feature_factory_dict[column] = {
"CountEncoder": CountEncoder(column=list(column)),
"TargetEncoder": TargetEncoder(column=list(column))
}
feature_factory_manager = FeatureFactoryManager(
feature_factory_dict=feature_factory_dict,
logger=logger,
split_num=split_num)
df = feature_factory_manager.all_predict(df)
os.makedirs(output_dir, exist_ok=True)
params = {
'objective': 'binary',
'num_leaves': 32,
'min_data_in_leaf': 15, # 42,
'max_depth': -1,
'learning_rate': 0.3,
'boosting': 'gbdt',
'bagging_fraction': 0.7, # 0.5,
'feature_fraction': 0.9,
'bagging_seed': 0,
'reg_alpha': 5, # 1.728910519108444,
'reg_lambda': 5,
def main(params: dict, output_dir: str):
import mlflow
print("start params={}".format(params))
model_id = "train_0"
logger = get_logger()
# df = pd.read_pickle("../input/riiid-test-answer-prediction/train_merged.pickle")
df = pd.read_pickle(
"../input/riiid-test-answer-prediction/split10/train_0.pickle"
).sort_values(["user_id", "timestamp"]).reset_index(drop=True)
if is_debug:
df = df.head(30000)
df["prior_question_had_explanation"] = df[
"prior_question_had_explanation"].fillna(-1)
column_config = {
("content_id", "content_type_id"): {
"type": "category"
},
"user_answer": {
"type": "leakage_feature"
},
"part": {
"type": "category"
},
"prior_question_elapsed_time_bin300": {
"type": "category"
},
"duration_previous_content_bin300": {
"type": "category"
},
"prior_question_had_explanation": {
"type": "category"
},
"rating_diff_content_user_id": {
"type": "numeric"
},
"qq_table2_mean": {
"type": "numeric"
},
"qq_table2_min": {
"type": "numeric"
}
}
if not load_pickle or is_debug:
feature_factory_dict = {"user_id": {}}
feature_factory_dict["user_id"][
"DurationPreviousContent"] = DurationPreviousContent()
feature_factory_dict["user_id"][
"ElapsedTimeBinningEncoder"] = ElapsedTimeBinningEncoder()
feature_factory_dict["user_id"]["PreviousAnswer2"] = PreviousAnswer2(
groupby="user_id",
column="content_id",
is_debug=is_debug,
model_id=model_id,
n=300)
feature_factory_dict["user_id"][
"UserContentRateEncoder"] = UserContentRateEncoder(
rate_func="elo", column="user_id")
feature_factory_dict["user_id"]["QuestionQuestionTableEncoder2"] = \
QuestionQuestionTableEncoder2(
model_id=model_id,
is_debug=is_debug,
past_n=100,
min_size=300
)
feature_factory_manager = FeatureFactoryManager(
feature_factory_dict=feature_factory_dict,
logger=logger,
split_num=1,
model_id="train_0",
load_feature=not is_debug,
save_feature=not is_debug)
print("all_predict")
df = feature_factory_manager.all_predict(df)
df = df[[
"user_id", "content_id", "content_type_id", "part", "user_answer",
"answered_correctly", "prior_question_elapsed_time_bin300",
"duration_previous_content_bin300",
"prior_question_had_explanation", "rating_diff_content_user_id",
"qq_table2_mean", "qq_table2_min"
]].replace(-99, -1)
df["qq_table2_mean"] = df["qq_table2_mean"].fillna(0.65)
df["qq_table2_min"] = df["qq_table2_min"].fillna(0.6)
print(df.head(10))
print("data preprocess")
train_idx = []
val_idx = []
np.random.seed(0)
for _, w_df in df[df["content_type_id"] == 0].groupby("user_id"):
if np.random.random() < 0.01:
# all val
val_idx.extend(w_df.index.tolist())
else:
train_num = int(len(w_df) * 0.95)
train_idx.extend(w_df[:train_num].index.tolist())
val_idx.extend(w_df[train_num:].index.tolist())
ff_for_transformer = FeatureFactoryForTransformer(
column_config=column_config,
dict_path="../feature_engineering/",
sequence_length=params["max_seq"],
logger=logger)
ff_for_transformer.make_dict(df=df)
n_skill = len(ff_for_transformer.embbed_dict[("content_id",
"content_type_id")])
if not load_pickle or is_debug:
df["is_val"] = 0
df["is_val"].loc[val_idx] = 1
w_df = df[df["is_val"] == 0]
w_df["group"] = (
w_df.groupby("user_id")["user_id"].transform("count") -
w_df.groupby("user_id").cumcount()) // params["max_seq"]
w_df["user_id"] = w_df["user_id"].astype(
str) + "_" + w_df["group"].astype(str)
group = ff_for_transformer.all_predict(w_df)
dataset_train = SAKTDataset(group,
n_skill=n_skill,
max_seq=params["max_seq"])
del w_df
gc.collect()
ff_for_transformer = FeatureFactoryForTransformer(
column_config=column_config,
dict_path="../feature_engineering/",
sequence_length=params["max_seq"],
logger=logger)
if not load_pickle or is_debug:
group = ff_for_transformer.all_predict(df[df["content_type_id"] == 0])
dataset_val = SAKTDataset(group,
is_test=True,
n_skill=n_skill,
max_seq=params["max_seq"])
os.makedirs("../input/feature_engineering/model080", exist_ok=True)
if not is_debug and not load_pickle:
with open(f"../input/feature_engineering/model080/train.pickle",
"wb") as f:
pickle.dump(dataset_train, f)
with open(f"../input/feature_engineering/model080/val.pickle",
"wb") as f:
pickle.dump(dataset_val, f)
if not is_debug and load_pickle:
with open(f"../input/feature_engineering/model080/train.pickle",
"rb") as f:
dataset_train = pickle.load(f)
with open(f"../input/feature_engineering/model080/val.pickle",
"rb") as f:
dataset_val = pickle.load(f)
print("loaded!")
dataloader_train = DataLoader(dataset_train,
batch_size=params["batch_size"],
shuffle=True,
num_workers=1)
dataloader_val = DataLoader(dataset_val,
batch_size=params["batch_size"],
shuffle=False,
num_workers=1)
model = SAKTModel(n_skill,
embed_dim=params["embed_dim"],
max_seq=params["max_seq"],
dropout=dropout)
param_optimizer = list(model.named_parameters())
no_decay = ['bias', 'LayerNorm.bias', 'LayerNorm.weight']
optimizer_grouped_parameters = [{
'params':
[p for n, p in param_optimizer if not any(nd in n for nd in no_decay)],
'weight_decay':
0.01
}, {
'params':
[p for n, p in param_optimizer if any(nd in n for nd in no_decay)],
'weight_decay':
0.0
}]
optimizer = AdamW(
optimizer_grouped_parameters,
lr=params["lr"],
weight_decay=0.01,
)
num_train_optimization_steps = int(len(dataloader_train) * epochs)
scheduler = get_linear_schedule_with_warmup(
optimizer,
num_warmup_steps=params["num_warmup_steps"],
num_training_steps=num_train_optimization_steps)
criterion = nn.BCEWithLogitsLoss()
model.to(device)
criterion.to(device)
for epoch in range(epochs):
loss, acc, auc, auc_val = train_epoch(model, dataloader_train,
dataloader_val, optimizer,
criterion, scheduler, device)
print("epoch - {} train_loss - {:.3f} auc - {:.4f} auc-val: {:.4f}".
format(epoch, loss, auc, auc_val))
preds = []
labels = []
with torch.no_grad():
for item in tqdm(dataloader_val):
x = item["x"].to(device).long()
target_id = item["target_id"].to(device).long()
part = item["part"].to(device).long()
label = item["label"].to(device).float()
elapsed_time = item["elapsed_time"].to(device).long()
duration_previous_content = item["duration_previous_content"].to(
device).long()
prior_question_had_explanation = item["prior_q"].to(device).long()
user_answer = item["user_answer"].to(device).long()
rate_diff = item["rate_diff"].to(device).float()
qq_table_mean = item["qq_table_mean"].to(device).float()
qq_table_min = item["qq_table_min"].to(device).float()
output = model(x, target_id, part, elapsed_time,
duration_previous_content,
prior_question_had_explanation, user_answer,
rate_diff, qq_table_mean, qq_table_min)
preds.extend(torch.nn.Sigmoid()(
output[:, -1]).view(-1).data.cpu().numpy().tolist())
labels.extend(label[:, -1].view(-1).data.cpu().numpy().tolist())
auc_transformer = roc_auc_score(labels, preds)
print("single transformer: {:.4f}".format(auc_transformer))
df_oof = pd.DataFrame()
# df_oof["row_id"] = df.loc[val_idx].index
print(len(dataloader_val))
print(len(preds))
df_oof["predict"] = preds
df_oof["target"] = labels
df_oof.to_csv(f"{output_dir}/transformers1.csv", index=False)
"""
df_oof2 = pd.read_csv("../output/ex_237/20201213110353/oof_train_0_lgbm.csv")
df_oof2.columns = ["row_id", "predict_lgbm", "target"]
df_oof2 = pd.merge(df_oof, df_oof2, how="inner")
auc_lgbm = roc_auc_score(df_oof2["target"].values, df_oof2["predict_lgbm"].values)
print("lgbm: {:.4f}".format(auc_lgbm))
print("ensemble")
max_auc = 0
max_nn_ratio = 0
for r in np.arange(0, 1.05, 0.05):
auc = roc_auc_score(df_oof2["target"].values, df_oof2["predict_lgbm"].values*(1-r) + df_oof2["predict"].values*r)
print("[nn_ratio: {:.2f}] AUC: {:.4f}".format(r, auc))
if max_auc < auc:
max_auc = auc
max_nn_ratio = r
print(len(df_oof2))
"""
if not is_debug:
mlflow.start_run(experiment_id=10, run_name=os.path.basename(__file__))
for key, value in params.items():
mlflow.log_param(key, value)
mlflow.log_metric("auc_val", auc_transformer)
mlflow.end_run()
torch.save(model.state_dict(), f"{output_dir}/transformers.pth")
del model
torch.cuda.empty_cache()
with open(f"{output_dir}/transformer_param.json", "w") as f:
json.dump(params, f)
if is_make_feature_factory:
# feature factory
feature_factory_dict = {"user_id": {}}
feature_factory_dict["user_id"][
"DurationPreviousContent"] = DurationPreviousContent(
is_partial_fit=True)
feature_factory_dict["user_id"][
"ElapsedTimeBinningEncoder"] = ElapsedTimeBinningEncoder()
feature_factory_manager = FeatureFactoryManager(
feature_factory_dict=feature_factory_dict,
logger=logger,
split_num=1,
model_id="all",
load_feature=not is_debug,
save_feature=not is_debug)
ff_for_transformer = FeatureFactoryForTransformer(
column_config=column_config,
dict_path="../feature_engineering/",
sequence_length=params["max_seq"],
logger=logger)
df = pd.read_pickle(
"../input/riiid-test-answer-prediction/train_merged.pickle")
if is_debug:
df = df.head(10000)
df = df.sort_values(["user_id", "timestamp"]).reset_index(drop=True)
feature_factory_manager.fit(df)
df = feature_factory_manager.all_predict(df)
for dicts in feature_factory_manager.feature_factory_dict.values():
for factory in dicts.values():
factory.logger = None
feature_factory_manager.logger = None
with open(f"{output_dir}/feature_factory_manager.pickle", "wb") as f:
pickle.dump(feature_factory_manager, f)
ff_for_transformer.fit(df)
ff_for_transformer.logger = None
with open(
f"{output_dir}/feature_factory_manager_for_transformer.pickle",
"wb") as f:
pickle.dump(ff_for_transformer, f)
def make_feature_factory_manager(split_num, model_id=None):
logger = get_logger()
feature_factory_dict = {}
for column in ["user_id", "content_id"]:
is_partial_fit = (column == "content_id" or column == "user_id")
if type(column) == str:
feature_factory_dict[column] = {
"TargetEncoder": TargetEncoder(column=column, is_partial_fit=is_partial_fit)
}
else:
feature_factory_dict[column] = {
"TargetEncoder": TargetEncoder(column=list(column), is_partial_fit=is_partial_fit)
}
feature_factory_dict["user_id"]["DurationPreviousContent"] = DurationPreviousContent(is_partial_fit=True)
feature_factory_dict["user_id"]["PastNTimestampEncoder"] = PastNFeatureEncoder(column="timestamp",
past_ns=[2, 3, 4, 5],
agg_funcs=["vslast"],
remove_now=False)
feature_factory_dict["user_id"]["StudyTermEncoder2"] = StudyTermEncoder2(is_partial_fit=True)
feature_factory_dict["user_id"]["ElapsedTimeMeanByContentIdEncoder"] = ElapsedTimeMeanByContentIdEncoder()
feature_factory_dict["user_id"]["CountEncoder"] = CountEncoder(column="user_id", is_partial_fit=True)
feature_factory_dict[("user_id", "part")] = {
"UserContentRateEncoder": UserContentRateEncoder(column=["user_id", "part"],
rate_func="elo")
}
for column in ["user_id", "content_id", "part", ("user_id", "part"), ("user_id", "prior_question_had_explanation")]:
if column not in feature_factory_dict:
feature_factory_dict[column] = {}
if type(column) == str:
feature_factory_dict[column][f"MeanAggregatorShiftDiffTimeElapsedTimeby{column}"] = MeanAggregator(column=column,
agg_column="duration_previous_content_cap100k",
remove_now=False)
feature_factory_dict[column][f"MeanAggregatorStudyTimeby{column}"] = MeanAggregator(column=column,
agg_column="study_time",
remove_now=False)
else:
feature_factory_dict[column][f"MeanAggregatorShiftDiffTimeElapsedTimeby{column}"] = MeanAggregator(column=list(column),
agg_column="duration_previous_content_cap100k",
remove_now=False)
feature_factory_dict[column][f"MeanAggregatorStudyTimeby{column}"] = MeanAggregator(column=list(column),
agg_column="study_time",
remove_now=False)
feature_factory_dict["user_id"]["CategoryLevelEncoderPart"] = CategoryLevelEncoder(groupby_column="user_id",
agg_column="part",
categories=[2, 5])
feature_factory_dict["user_id"]["UserContentNowRateEncoder"] = UserContentNowRateEncoder(column="part",
target=[1, 2, 3, 4, 5, 6, 7],
rate_func="elo")
feature_factory_dict["user_id"]["PreviousAnswer2"] = PreviousAnswer2(groupby="user_id",
column="content_id",
is_debug=is_debug,
model_id=model_id,
n=300)
feature_factory_dict["user_id"]["PreviousNAnsweredCorrectly"] = PreviousNAnsweredCorrectly(n=5,
is_partial_fit=True)
feature_factory_dict[f"previous_5_ans"] = {
"TargetEncoder": TargetEncoder(column="previous_5_ans")
}
feature_factory_dict["user_id"]["QuestionLectureTableEncoder2"] = QuestionLectureTableEncoder2(model_id=model_id,
is_debug=is_debug,
past_n=100,
min_size=300)
feature_factory_dict["user_id"]["QuestionQuestionTableEncoder2"] = QuestionQuestionTableEncoder2(model_id=model_id,
is_debug=is_debug,
past_n=100,
min_size=300)
feature_factory_dict["user_id"]["UserContentRateEncoder"] = UserContentRateEncoder(column="user_id",
rate_func="elo")
feature_factory_dict["user_id"]["PreviousContentAnswerTargetEncoder"] = PreviousContentAnswerTargetEncoder(min_size=300)
feature_factory_manager = FeatureFactoryManager(feature_factory_dict=feature_factory_dict,
logger=logger,
split_num=split_num,
model_id=model_id,
load_feature=not is_debug,
save_feature=not is_debug)
return feature_factory_manager