v2.b579 (Version 2, Build 579) - Released February 05, 2018
The new release contains the following new features and bug fixes.
Plate Transfer Interface:
Riffyn's Plate Transfer interface can be used to record transfers between microtiter plates. Experiments are frequently initiated by carefully mapping input resources to specific wells of an initial set of plates. Riffyn’s plate transfer interface will assist in tracking subsequent transfer events between plates. A range of plate sizes (6-well, 24-well, 96-well, 384-well) and standard plate-to-plate transfers are supported. The plate transfer interface is accessible from the “Plate” tab in Measure Mode.
Note - New system fixed properties have been added which aid automation in the new plate transfer functionality. They are "plate ID", "row", "column", "well". If your organization has created custom property types with the same names it is suggested that their use be discontinued.
In-app formulas support a wide range of excel-like functions within measure mode. Prior to this release it was necessary to calculate values either on-export from Riffyn or in other applications such as Excel. The new in-app formulas can use data from and write data to both plan and actual values. Formulas can be defined in design mode by editing resource types or properties, or in measure modes by right-clicking on the column name. Numeric or character values can be written to properties or resource names in measure mode.
Riffyn’s new sequence generator allows run names, output resource names, and single value properties to be created for selected runs following a defined pattern in measure mode.
The export feature provides a way to define multiple configurations for exporting data from Riffyn. A configuration can use existing clean data procedures and join matching single and multi-valued data points on connected runs. This means that data can be exported by taking into consideration both the run connections and the values of properties on the steps. This can yield a more concise data set.
Note, the current joining algorithm uses an inner join and exact value matching. This means that non-matching data rows will be dropped (not exported). In upcoming releases we will (1) update the joining logic to an outer join to preserve non-matching data rows, and (2) add capabilities for 'fuzzy' matching of data values.
The following new columns have been added to all exported data:
- run number - the run number as a numerical value
- unit - will contain mixed units if property units vary from experiment to experiment.
The following columns have been updated:
- run description changed to run name - to match the Riffyn user interface.
- resource label has changed to resource rifID - to better reflect the contents of the column
- "spec | valid-options" label on character properties has changed to "spec | target" for consistency with the UI labels.
- event id column has been updated to report the following values:
- For single-valued data (per run) the event id will show "None-0"
- If a data cleaning procedure has been applied, the event id will show "CLUSTER:M-N" where M is group/cluster number and N is nth row in the cluster.
The following improvements and new features have been added to the Riffyn API:
- New Endpoints:
- deleteResource, delete a specific resource (the resource will remain in experiments where is has been used but will no longer be available for new experiments)
- getRun, you can now get a specific run directly for an experiment.
- updateRunStatus, start, stop, mark as complete, reset, and restart a run.
- Endpoints to determines if a user has access to an entity:
- Updated Endpoints:
- ListResources endpoint can now be filtered by:
- experiment Id
- listTasks endpoint: added these filters
- GetExperimentDataRaw and getExperimentDataClean endpoints: removed repeated Header MetaData in the datatables and eliminated RunMap
- setResource endpoint has been deprecated - should now use the assignInput endpoint. Functionality has not changed
- getExperimentStatus name changed to getExperimentExportStatus for clarity. Functionality stays the same
- ListResources endpoint can now be filtered by:
RiffynTools JMP AddIn:
The following improvements and new features have been added to the RiffynTools JMP AddIn:
- Process and experiment browser for fetching data sets
- Advanced experiment join. includes a fully automated joining algorithm; it handles complex topologies, supports both 'AND' and 'OR' joining logic when there are multiple connecting resources between processes. Allows joining on resource name or id.
- Visual data tagging tool. Select data points on any graph in JMP generated from a Riffyn data table, and write any data tag to those points. Useful for hit-picking and data quality tagging (e.g., excluding outliers).
- DoE tool. Tool for constructing DoEs on Riffyn experiment using JMP Custom DoE platform and then saving those designs directly from JMP to the plan on Riffyn experiments.
- Reorganized toolbox menus.
- Upgraded for compatibility with JMP 14, Windows, and Riffyn SDE 2.0. Eliminates dependency on curl for JMP 14 and above (with some performance benefits). Provided curl installer for JMP 13 users.
- Unit annotation utility (add units, if missing from data table) and append them to column properties or names).
- Column abbreviation tools. Utility for shortening/encoding portions of column names including process names, step names and property column headings.
- Improved column ordering.
- Column merge utility. Combines properties that get split into separate columns in the data table when properties are replaced or move to other steps between process versions.
Issues Fixed in Measure Mode
- Eyedropper/search icons and updated tooltip will prompt users to leave aggregated view when trying to assign resources
- Aggregation displays an appropriate status when one of its resource assignments is incorrect
- Number of runs selected in aggregates is displayed instead of the number of aggregates
- Changing the run type, start/stop, or a resource name will mark an experiment as stale
- Clicking a multi-valued-data cell will select the run (in addition to opening the multi-valued data editor if not already open)
- The enter key or mouse-click on the search icon will trigger a search in the top panel if text has been entered when searching for a resource
- Shift-select will use the run sort order not the run number when selecting runs
- Next/Prev run buttons in multi-valued data editor move to the next/previous runs when only 1 run initially selected
- All runs are deselected by clicking on the white space below the runs.
- Run selection does not wrap to top run on hitting enter key in the last run
- Resource assignment panel is closed when editing a property
- Purpose added to experiment list table