etna
etna copied to clipboard
tinkoff-ai
ETNA – Time-Series Library
### 🚀 Feature Request As the authors of [statsforecast](https://github.com/Nixtla/statsforecast#maintenance-FedericoGarza) state their auto_arima is 20x faster than the version form pmdarima(which we currently use).There for it might be useful to add...
### 🚀 Feature Request We decided that we want to enable passing context into `forecast` method of the model. For this we should for each model define its context size....
### 🚀 Feature Request Experiments show that linear regression with positive coefficients shows better results comparing with no restrictions to the coefficients. So it is better to change the default...
## Before submitting (must do checklist) - [x] Did you read the [contribution guide](https://github.com/tinkoff-ai/etna/blob/master/CONTRIBUTING.md)? - [x] Did you update the docs? We use Numpy format for all the methods and...
### 🚀 Feature Request Add `MLPModel` model for etna deep model family ### Proposal * You can use [as ref](https://github.com/tinkoff-ai/etna/blob/master/etna/models/nn/rnn.py) ```python class MLPNet(DeepBaseNet): def __init__( self, input_size: int, hidden_size: List[int],...
### 🚀 Feature Request Create metric, which counts the percent of times, when the entire metric was smaller that threshold. ### Proposal - In file `metrics/metric.py` create class: ``` class...
### 🚀 Feature Request Create metric, which use the custom penalty for over- and under- forecasting. This might be useful when the over- and under- forecasting is harmful differently. ###...
### 🚀 Feature Request Add method `predict` to pipelines. ### Motivation Introduce more flexible ways of forecasting to our architecture. ### Proposal This task is blocked by #783. Add a...
### 🚀 Feature Request Create notebook about forecasting strategies in etna ### Proposal - Create notebook about the forecasting strategies in etna, it should contain examples with: 1. Pipeline 2....
### Preposition: We have multiple models and feature generators to use. We don't know what model or features to choose from. We have limited cpu and time resources. We have...