Within MoveApps, existing Apps can be combined into a Workflow, allowing you to define an ordered set of steps to access, process and analyse data. These Workflows can then be saved, edited and used for specific use cases. Before starting, we recommend to check out our video tutorials and to follow along with the example Workflow Hello world!.
Every Workflow starts with an App that loads data into the system (e.g. from Movebank, the user's local system, Dropbox/Google Drive or an output from another MoveApps workflow). The data are then passed on to the next App and analysed by it accordingly. Every App requires data of a defined input type (mostly move2::move2_loc, but others are available and the numbers are growing) and can pass on output of a defined type. See Translator Apps for instructions on how to connect Apps with different output and input types.
Some Apps provide an interactive User Interface (R-Shiny). Many Apps produce output products or artifacts, which can be downloaded as files, in formats such as .pdf, .csv, .png, .html among others. App output data can also be downloaded as an object of the defined App output type in R format app-output.rds or Python format app-output.pickle. Any App artifacts or outputs are also available via our API.
From the main menu, select Workflows and then Create New Workflow. After providing a name for the Workflow (also consider adding a Workflow Category), you will be asked to Select A Data Source. Now choose to retrieve location or non-location data from Movebank or another MoveApps workflow (Workflow Product Retriever), or location data (!) from your local system or your personal cloud storage on Google Drive or Dropbox. For the latter option the data have to be provided as an R object of class 'move2::move2' saved in .rds format or as .csv containing information about timestamp, x, y coordinates and track ID, preferably in a Movebank format (see Details). All Data Sources require that you specify parameters and settings. After providing these, the Workflow area with the initial App appears . Note that additional Data Sources can be downloaded by adding e.g. another Movebank Location App (see below for instructions on adding Apps). All data sets will be combined, thus allowing the joined analysis of data from e.g. different Movebank studies.
Note that recently, we have added the option to download non-location data from Movebank. These are e.g. accelerometer, accessory measurement, heart rate or geolocation raw data. In the Movebank Non-Location App, beware of the option to select only studies that contain non-location data. Furthermore, it is only allowed to download one type of non-location data at a time. Look out for Apps that work with those data types in the App Browser; they should have the Input/Output type "move2::move2_nonloc".
By clicking on the + to the right of an App container, you can browse, search and select the next App to add to your Workflow for filtering, analysis or visualisation of the downloaded data. You can only insert Apps within the Workflow if they are compatible with the required input and output (IO) types. The list of Apps to choose from will only include those that comply with the required IO types for the specific position in your Workflow. To see also incompatible Apps activate the Include incompatible Apps switch. Use the Translator Apps to transform IO types enabling to connect Apps with different IO types.
By clicking on the App name you are directed to the App Details that also contain links to the documentation, code and issues.
!> Note that depredated Apps do not appear in this list by default. Deprecated Apps cannot be added to new Workflows any more, because the App developer has stopped maintenance.
Select Start Workflow to begin running the Apps within a Workflow in the order that you have arranged them. You can follow the progress of the Workflow by the change in colour of the action point in the top right of each App container. On the right side of the Workflow name there is also an action point which will indicate if the Workflow Instance is "running" or "stopped". Note that if the Workflow contains a Shiny App its status will be "running" for a few hours, also after finishing, in order to allow the interaction with the App User Interface.
The Workflow continues running and results are stored even if you leave the site moveapps.org. Any results or artefacts/products files can be looked up and downloaded later. Select Rerun or Stop Workflow to interrupt the run of the Workflow. Each Workflow can also be scheduled to run regularly in an automatic mode at a fixed date and time. This option (Schedule Instance) can be selected in the menu next to Output.
When running a Workflow, Apps may give an error and subsequent Apps will not run.
We recommend the following actions to try to solve the error.
Take a look at the error message by clicking on Error details. The error message may already indicate the problem.
In some cases, the Error details do not provide helpful information. In these cases, the app logs may contain more information about the error. The App Logs can be accessed by clicking on app logs in the window with error details, or through the App Menu (see Menu, Settings and Logs).
In the App documentation (click on the book icon in the App container), there often is a section covering the most common errors. Your error may be covered here together with a potential solution.
In some cases, the error details or App logs indicate a problem with the App settings. In these cases, and also generally when encountering an error, we recommend to revise the App settings. App settings can be accessed through the App Menu (see Menu, Settings and Logs), or for Shiny Apps through the User Interface.
App users can create an "issue" in the App's GitHub repository to notify the App developer about problems with their App. You can access the issue site by clicking on issue at the App repository in the window with error details, or through the App Details that can be accessed via the App Menu (see Menu, Settings and Logs). Have a look at the list of issues on the App's repository, maybe other users encountered the same problem.
If not, create a New issue in GitHub in the App repository (you need a GitHub account for this). To facilitate solving the error we recommend attaching to the issue the log file of the App. This file, log.txt, can be found in the output of each App. In the case of interactive Apps (Shiny Apps) this file only becomes available once the App in inactive (after 8h), but can always be downloaded from the logs overview which can be accessed via Show Logs in the App Menu (see Menu, Settings and Logs). Before making the logs public, please ensure that it does not contain any sensitive information.
If you do not have a GitHub account, you can always also email the App developer. The email can be found in the the App Details that can be accessed via the App Menu (see Menu, Settings and Logs). To speed up the error solving remember to attach the log.txt, that can be found in the output of each App. In the case of interactive Apps (Shiny Apps) this file only becomes available once the App in inactive (after 8h), but can always be downloaded from the logs overview which can be accessed via Show Logs in the App Menu (see Menu, Settings and Logs). Before sharing the logs, please ensure that it does not contain any sensitive information.
You can always contact MoveApps support (support@moveapps.org) for any issue you may encounter.
When an error occurs, by clicking on MoveApps support you can directly send a message to the support (with you in CC) that will contain your email, the name of the App and the logs of the App.
If the App has issues but is not producing an error, you can always download log.txt, that can be found in the output of each App. In the case of interactive Apps (Shiny Apps) this file only becomes available once the App in inactive (after 8h), but can always be downloaded from the logs overview which can be accessed via Show Logs in the App Menu (see Menu, Settings and Logs). Before sharing the logs, please ensure that it does not contain any sensitive information.
There is a menu on the right of each App container in a Workflow. The R and Python apps include the below options. R-Shiny-Apps do not provide Settings in this menu, but directly in the User Interface (UI). Those settings can be stored on MoveApps from within the UI (bookmark feature).
Settings: View or change App parameters.App Details: View all information and details about the App.Pin to this app: By pinning the Workflow to an App, you will retain the results of this App and all Apps preceding it in the Workflow, so that only subsequent Apps are re-executed when you re-run the Workflow. The App you pinned and all preceding Apps are underlaid in grey. Note that Shiny-Apps close the running UI if pinned, but retain the output data. ThePincan be removed any time by clicking onRemove pinbelow in the greyed out area.Show Logs: View the data protocols, which can contain important information if you receive errors or unexpected results.Update App: If a new version of the App is available it can be updated here. A warning icon in the App container will notify of available App updates.Delete: Remove the App from your Workflow.
!> Note that there is a small book icon in the top left corner of each App container, leading directly to the Github App documentation, where the App developer has provided helpful information that describes when and how the App can be used and how common errors can be solved.
Each App that returns data creates a summary of the output data (Cargo Agent), which can be accessed via the green info button that appears on the right side of the App container. This summary depends on the output type of the App. For example, for move2::move2_loc it includes, among others, animal attributes (see Movebank for definitions) the bounding box, projection, sensor types and time range of the data, as well as the total number positions (positions_total_number). Depending on the output type of an App the summary details can differ. Note the section "Unexpected Results?" at the bottom that provides helpful comments if errors occur.
For R-Shiny Apps that return User Interfaces (UI), the UI can be accessed after the run of the Workflow via Open App UI. There, you can examine results and edit settings, depending on the options that the App developer has programmed. On the bottom left corner of each Shiny App the button Store settings will store the current settings of the App for future workflow runs. When Apps are deleted or added to the workflow, all subsequent Apps will reset to the default values. Apps for which settings have been stored will contain an item called Stored Settings in the App Outputs.
In addition to the direct click-download of files that appear in the App containers after the workflow has run, the Output button gives you access to all output files that the Workflow is generating. Each output file of the Apps can be downloaded locally or uploaded to a cloud storage provider. A selection of files can be bundled and downloaded as a zip file. In case of one or more .pdf products, it is also possible to download a bundled pdf, unsecured or as encrypted file for sensitive tracking data (see buttons on top right).
At the top of the page, you can access our API service that allows to create a stable http link to any of the products created by the workflow.
In the Wokrflow output we provide a file named "Selected App-Settings.pdf", this file will contain the selected settings for each App of the Workflow instance. In order to save the settings of interactive Apps with a User Interface (UI), the Store settings button must be clicked after customizing the settings.
Each Workflow can be run and saved in several Instances, allowing you to combine the same Apps in the same order, but with different settings. It is important to note that changes in the selection and order of Apps (not their settings) in one Workflow Instance will be applied to all other Instances of the Workflow.
Every additional Workflow Instance can be added via the left side menu via + New instance or can be added in the Your Workflows overview (via the main menu in your Dashboard). You can find the button + New instance next to your existing instances.
In the menu of each Instance (three dots next to the Output button), it is possible to Edit Instance Details, Edit Workflow Details and Schedule Instance (see scheduling) or to Delete Workflow Instance. If not empty, the description/details of the Workflow and/or Workflow Instance will be shown on the top of the page.
Once an Instance of a Workflow is opened, you can switch between all Instances of the Workflow through the main menu on the top left side. To select another Workflow, use the button Workflows.
Workflows can be grouped by assigning them a Workflow Category name. If no Category is assigned, they will appear under "Uncategorized" (2). At any time Categories(1), Workflows(5) and Instances(4) can be edited, and workflows can be shared or published (5) via the Your Workflows overview page. Also nwe Categories can be added at any time (3). When adding Workflows that have been privately shared with you, they appear in the Category "Imported Privately Shared Workflows". Added public Workflows are initially stored in the Category "Imported Public Shared Workflows". Any added Workflows are only copies, they can be changed and reworked in any way and the Workflow Category adapted.
Search: you can search in the title and description of your Workflows and Workflow instances.
Categories: display of all categories. You can select one category at a time, edit them via the menu (1) or add a new category. The number next to each category name is the number of Workflows that have this category assigned. If non are selected, all Workflows will be displayed in the table below.
Table: you can use each column to order your workflows. If Workflow instance names are long, it might happen that the table does not fit into your window size and seems to be cut off, you can use bar at the bottom of the table to move the table left and right, or reduce the zoom in your browser to see the entire table.
-
Workflow: in this column all Workflows with their respective Workflow instances are displayed. By clicking on a Workflow instance name you will open it. Each Workflow instance can be edited via the menu (2). At the bottom of the table (3) you can select if the instances within each Workflow should be ordered alphabetically or by 'last start', i.e. how long ago you last clicked the button
StartorRerun, or a scheduled instance started. On the right end of the Workflow column you can click (4) to see the Workflow and Workflow instance description. The menu (5) to edit and share the Workflow can be found at the right side of the table. -
Category: in this column the categories are displayed, select a category by clicking on it.
-
Last start: this column displays, for any instance of the Workflow, how long ago you last clicked the button
StartorRerun, or a scheduled instance started. -
Created: this column displays how long ago the Workflow was created.