Importing Training Data

If you have a set of trades in a CSV file or other format you can easily import these to train your strategy rather than entering training points manually.

To use the import feature, choose the strategy from your strategy list, or create a new one by following this tutorial.

From the chart page, switch on training mode by clicking the hat icon in the toolbar. Switch to the Upload Data tab. Copy and paste your data into the text box. Each line should be correctly formatted, with the trade open time stamp, close time stamp, and the side.

Card image cap

Importing training data

After pasting your data into the text box, the system will try to detect the field separator. The separator is shown in the Field separator box. Check that this matches your imported data.

The time stamp for each open and close should fall within valid bars of your chosen data series. If your trade time stamps are not in UTC (Universal Time Coordinates), enter the offset in hours of your data to UTC so that bars are matched correctly.

Card image cap

Import results

After completing the above, click the Import button to start the import. The import will tell you how many points were successfully imported, and how many were ignored due to errors. A point might be ignored, if there is no bar in the data series that matches either the open or the close. If you see a lot of these types of errors it might be because the correct UTC offset hasn’t been set.

Card image cap

List of imported points

Once the import is complete, you can switch to the Training Points tab to review the data. You can also check the points visually on the chart. To train your strategy with the new training set, click the Train button in the bottom menu.

Hints for Imported Training Data

Machine learning systems can be over trained or under trained. Therefore when importing training data, remember to follow best practices listed here. If you have an existing strategy, it isn’t necessary or advisable to import every single trade as a training case. This can create an over trained system that won’t be able to generalize well to future price data.

Risk Disclosure: Information given on this website is for general purposes only and should not be construed as investment advice. Certain instruments shown here are complex and may come with a high risk of losing money rapidly due to leverage. You should consider whether you understand how these instruments work and whether you can afford to take the high risk of losing your money. Market data displayed is indicative and is not a solicitation to buy or sell.