pykt.datasets package

Submodules

pykt.datasets.data_loader module

class pykt.datasets.data_loader.KTDataset(file_path, input_type, folds, qtest=False)

Bases: Dataset

Dataset for KT

can use to init dataset for: (for models except dkt_forget)

train data, valid data common test data(concept level evaluation), real educational scenario test data(question level evaluation).

Parameters

• file_path (str) – train_valid/test file path

• input_type (list[str]) – the input type of the dataset, values are in [“questions”, “concepts”]

• folds (set(int)) – the folds used to generate dataset, -1 for test data

• qtest (bool, optional) – is question evaluation or not. Defaults to False.

pykt.datasets.dkt_forget_dataloader module

class pykt.datasets.dkt_forget_dataloader.DktForgetDataset(file_path, input_type, folds, qtest=False)

Bases: Dataset

Dataset for dkt_forget

can use to init dataset for: dkt_forget

train data, valid data common test data(concept level evaluation), real educational scenario test data(question level evaluation).

Parameters

• file_path (str) – train_valid/test file path

• input_type (list[str]) – the input type of the dataset, values are in [“questions”, “concepts”]

• folds (set(int)) – the folds used to generate dataset, -1 for test data

• qtest (bool, optional) – is question evaluation or not. Defaults to False.

calC(row)

log2(t)

pykt.datasets.init_dataset module

pykt.datasets.init_dataset.init_dataset4train(dataset_name, model_name, data_config, i, batch_size)

pykt.datasets.init_dataset.init_test_datasets(data_config, model_name, batch_size)

pykt.datasets.init_dataset.update_gap(max_rgap, max_sgap, max_pcount, cur)

pykt.datasets.lpkt_dataloader module

class pykt.datasets.lpkt_dataloader.LPKTDataset(file_path, at2idx, it2idx, input_type, folds, qtest=False)

Bases: Dataset

pykt.datasets.lpkt_utils module

pykt.datasets.lpkt_utils.generate_time2idx(data_config)

pykt.datasets.que_data_loader module

class pykt.datasets.que_data_loader.KTQueDataset(file_path, input_type, folds, concept_num, max_concepts, qtest=False)

Bases: Dataset

Dataset for KT

can use to init dataset for: (for models except dkt_forget)

train data, valid data common test data(concept level evaluation), real educational scenario test data(question level evaluation).

Parameters

• file_path (str) – train_valid/test file path

• input_type (list[str]) – the input type of the dataset, values are in [“questions”, “concepts”]

• folds (set(int)) – the folds used to generate dataset, -1 for test data

• qtest (bool, optional) – is question evaluation or not. Defaults to False.

get_skill_multi_hot(this_skills)

Module contents