Skip to main content

Latest changes

June 20, 2024

New features

  • KPI Mapper: Support mapping to KPI proxies that are pre-calculated predictions.

Improvements

  • Company pages: Disabled refresh on tab focus to avoid losing selected time series in Data Explorer, and unsaved models or mappings in KPI Analyzer.
  • KPI Analyzer: Changed hit rate display to a ratio showing the number of "calls" (beat/miss) made, and the number of times a correct beat/miss call was made.
  • KPI Analyzer: Added descriptions of model types when training a custom model.

Bug fixes

  • KPI Analyzer: Fixed an issue where the KPI button could show a different KPI than the one actually picked when creating new KPI mappings.
  • Chart: Fixed an issue where a line could be missing from charts when copying or downloading if a custom color was not set.

June 13, 2024

New features

Improvements

  • KPI Analyzer: Added heatmap to columns "Final error", "Final hit rate", and "Predictions".
  • KPI Analyzer: Added tooltips to column headers with explanations of the column data.
  • KPI Analyzer: Support deleting any model with write access (not only own models).
  • Library: Added searching for alerts.
  • Library: Show KPI and company for KPI models, and change the Open link to open the model in KPI Analyzer.
  • Landing page: Persist selected watchlists/companies in the KPI Analyzer widget, and filter alerts for the same selection.

Bug fixes

  • KPI Analyzer: Fixed missing delete action for shared models where user has write-access.
  • DSL: Fixed an issue in combine_predictions where the calculated weights was wrong when the predictors had the same name.

June 6, 2024

New features

  • Landing page: The landing page has been revamped to include
    • KPI predictions
    • List of accessible dashboards
    • List of accessible data sets
    • Alerts feed
  • KPI Analyzer: It is now possible to share KPI models
  • KPI Analyzer: Added new column Delta/Error
  • KPI Analyzer: Added support for modelling types ARD Regression, Elastic Net, Elastic Net CV, Huber Regression, Linear Regression, and Ratio Prediction ML to custom KPI models.
  • KPI Mapper: Models will automatically be calculated as ratio models if the KPI timeseries is a ratio.
  • Model Builder: Added support for Ratio Prediction ML models.

Bug fixes

  • KPI Analyzer: Made sure that the first value for outside column filters are lower than the second value.

    Python SDK

  • New release of the Python SDK: Version 5.2.2.

May 30, 2024

New features

  • Alerts: Alerts are now available in the user's personal folder in the library, and can be shared with other users by moving to a shared folder, or by sharing your personal folder.

May 28, 2024

Improvements

Python SDK

May 21, 2024

Improvements

  • KPI Analyzer: Backtest revisions are now plotted for the selected period in the revisions chart.
  • KPI Analyzer: The mappings table can now be filtered and sorted on model accuracy, correlation, MAE and p-value.
  • Library: New and improved dialogs for creating, editing and sharing folders, and moving items between folders.

Bug fixes

  • Alpha Tester: Fixed an issue where some browser would not load more than 25 alpha tests when scrolling.
  • Signal Explorer: Disabled download button when only viewing non-exportable data, instead of giving a non-descriptive error message.
  • Signal Explorer/Library: Avoid overwriting signal expression when editing a signal in the Library view.

April 25, 2024

New features

  • KPI Analyzer: Added support for starring KPI mappings, in order to mark which KPIs you care most about, and a filter to view only starred KPI mappings. Starred mappings are sorted first in the KPI Analyzer.
  • KPI Analyzer: In the multi-company KPI Analyzer, added support for sorting and filtering KPIs by report date, prediction delta vs consensus, estimated error, and estimated hit rate. Also added filtering for only FactSet / Visible Alpha KPIs, or for specific KPIs.
  • KPI Analyzer: For each company-KPI, we now designate a primary model. By default, this will be your first custom model (in alphabetical order). You may change the primary model by selecting it from the dropdown next to the KPI in the KPI Analyzer monitoring view. The primary model is always used for sorting and filtering in the KPI Analyzer.

Improvements

  • KPI Analyzer: It is now possible to delete company-specific KPI mappings from the KPI Analyzer Mappings view.
  • KPI Analyzer: Display consensus estimated error, calculated on a like-for-like basis as each model.
  • KPI Analyzer: If an error occured while processing a KPI mapping, the error message is now available in the Mappings view as a tooltip in the "Final error" column.
  • Charts: Format tooltip with %/% pt/bps if the corresponding y-axis is formatted with these units.

Bug fixes

  • Import jobs: It is now possible to clear the alert email field.

April 11, 2024

New features

  • KPI Analyzer: We have added a multi-company page where it is possible to select multiple companies that you want to see KPI mappings and predictions for. From the main menu, click on "KPIs" -> "KPI Analyzer".
  • KPI Analyzer: We have added support for plotting how prediction errors change over time through a quarter, as data for the quarter is incrementally available. This is based on point-in-time backtesting.
  • Charts: New seasonality chart type to better visualize year-over-year changes in highly seasonal data.

Improvements

  • Data Explorer: When browsing the Data Explorer for a single company or entity, it is now possible to add a company or entity for comparison.
  • KPI Analyzer: Added support for copying a KPI mapping.
  • KPI Analyzer: Added support for deleting prediction models from the KPI Analyzer models page.

Bug fixes

  • Charts: Respect signal/signal group visibility toggle for pie, bar and column charts.
  • Dashboard: Removed some unnecessary refreshing and refetching of data for tables.

March 21, 2024

New features

  • Dashboards: We have made a new and simpler way to create and configure tables. It is now also possible to customize the number formatting (decimal places, units, suffixes) and tooltips of columns in the table.
  • KPI Analyzer: New estimated error and hit rate metrics are now available for each Auto & custom model, based on point-in-time backtesting. The previous error and hit rate metrics are still available but referred to as the "final" error and hit rate, as they are based on having all data available for each historical period.

Improvements

  • KPI Analyzer: In the models view, for each Auto & custom model, it is now possible to see details such as the user who created the model, when it was created, and when the predictions were last run.
  • KPI Analyzer: It is now possible to request a re-run of the Auto model. This may be useful to refresh the historical backtest error & hit rates, if many additional KPI mappings have been made. However, re-running the Auto model will reset the point-in-time model revision history; this is as if the Auto model was deleted and run from scratch.
  • KPI Analyzer: It is now possible to view consensus revisions for FactSet KPI estimates.
  • KPI Analyzer: Various small UI improvements.
  • KPI Mapper: The default resampling method is changed to sum.
  • Entities: When clicking on an entity in the app, we now display its entity type alongside its name.

Bug fixes

  • KPI picker: Fixed an issue where some KPIs were shown with duplicate entries.

March 7, 2024

New features

  • Entity overview: We have added an overview page for all types of entities, with links to Data Explorer and connected drill-downs.

Improvements

  • Company page: The company page header now includes a market data summary for the company.
  • Data Explorer: The Data Explorer can now be accessed in the scope of a single entity.
  • Drill-down pages: The drill-down pages have got a face lift with the same design as the company and entity overview pages.
  • KPI Analyzer: Show an error in input’s row in the model composition widget if a mapping has been excluded from a custom model.
  • KPI Analyzer: The probability widget now uses MAPE for non-ratio KPIs.
  • KPI Analyzer: It is not possible to click on mappings/models columns to navigate to the mapping/model views.
  • KPI Analyzer: The date of the last data point is now displayed in the mappings view.
  • KPI Analyzer: It is now possible to select companies as proxy entities.
  • Library: It is now possible to star your favourite charts.
  • Menu: Show recent updated models and strategies in the Modelling and Stragies menus, respectively.

Bug fixes

  • KPI Analyzer: Fixed an issue where the model closes when creating a new model, the user misses the checkbox and clicks the grey area around it instead.
  • KPI Analyzer: Fixed an issue where clicking on an entity chip in the popover toggles the selection of that row.