Overview
Datasets
Datasets are a fundamental entity in TimesDB, allowing users to input historical data for forecasting. Users can upload datasets onto the platform and associate tags to organize and categorize them.
Dataset Format
The dataset should be in CSV format with two columns: timestamp
and value
.
- timestamp: The timestamp should be in any parsable format, such as
YYYY-MM-DD HH:MM
orYYYY-MM-DD
. TimesDB supports various date and time formats. - value: The value column should contain numerical data. This can include integers or decimals, and both positive and negative values are supported.
Uploading Datasets
To upload a dataset, follow these steps:
- Navigate to the ‘Datasets’ section in the TimesDB platform.
- Click on ‘Upload Dataset’.
- Select the CSV file from your local system.
- Enter any relevant tags for categorization and organization.
- Click ‘Upload’ to complete the process.
Managing Datasets
After uploading, you can view, edit, and delete datasets in the ‘Datasets’ section. You can also add or remove tags as needed.
Dataset Best Practices
- Ensure your data is clean and preprocessed before uploading.
- Use descriptive tags for easy search and organization.
- Regularly update your datasets to maintain the accuracy of your forecasting models.