Torch-timeseries

func: adding a new model CATS
func: adding year into time features
func: add popular and set as default split, ETT for 6:2:2, others for 7:1:2
fix: default using only train data to scale

func: add iTransformer Forecast/Imputation/AnomalyDetection/UEAClassification
fix: fix a typo of Informer d_layer config

func: add Informer Forecast/Imputation/AnomalyDetection/UEAClassification
func: add Autoformer Forecast/Imputation/AnomalyDetection/UEAClassification
func: add FEDformer Forecast/Imputation/AnomalyDetection/UEAClassification
func: add PatchTST Forecast/Imputation/AnomalyDetection/UEAClassification

fix: fix a typo of the standard scaler
fix: fix some cli errors

func: Forecast/Imputation/UEAClassification AnomalyDetection experiments
func: pytexp entrypoint

fixed: dependencies issue of torch-timeserie pypi package