ORDS How-To 08: View, create, and share datasets

1. About this guide

This document is one of a series of How-To guides for the Online Research Database Service (ORDS). It will tell you how to view, create, and share ORDS datasets.

Please note that the ORDS is now in the process of being wound down. From 1 June 2017, the service will cease to be supported by the University of Oxford. This means that users should work on the assumption that ORDS datasets (and ORDS projects more generally) will cease to be available after this date. Please contact the Research Support team to discuss alternative options, or for assistance with data migration.

Contents

1.1. Pre-requisites

This How-To assumes that you have already registered to use the ORDS, and have either created or been given access to at least one project containing at least one database.

If this is not the case, please see the earlier guides in this series. These can be accessed via the main ORDS documentation page.

2. ORDS datasets

In ORDS terminology, a dataset is a set of saved query results.

2.1. Static and dynamic datasets

ORDS datasets can be static or dynamic.

  • A static dataset is a snapshot of the data at the time the dataset was created. If changes are subsequently made to the database, these will not affect the content of the dataset.
  • A dynamic dataset reflects the current contents of the database. If changes have been made to the database since the dataset was created, the dataset will update to include these.

Static datasets can be useful for preserving specific versions of data - for example, those on which research conclusions have been based. Dynamic datasets can be used as a way of re-running commonly used queries as part of ongoing data analysis.

For example, imagine a research project which is gathering data about pollution levels in rivers. Samples are regularly collected and analysed, and the results are added to an ORDS database. A dynamic dataset could be created to show key facts about pollution levels, as indicated by the most recent data: this might be viewed by project members on a daily or weekly basis. A series of static datasets could also be created (perhaps on a weekly or monthly basis) to show how the data has evolved over time. If a member of the project team presents some of their interim findings at a conference, a static dataset might be used to support these findings.

2.2. Dataset visibility

ORDS datasets have two visibility levels: public, and project members only.

If a dataset is made public, it will normally be accessible via the public ORDS project list. Public datasets have a stable URL, and thus it is possible to reference them in other research outputs, such as project websites and conference papers. (However, please see section 2.2.1. below about dataset availability.)

However, if a dataset forms part of a project which has been excluded from the public ORDS project list (see section 2.2.3. of How to create and manage projects for more details), it will not be publicly visible, even if this option is selected for that specific dataset. If the project is subsequently included in the public project list, the dataset will then automatically become publicly available. (This feature can be used to release a collection of datasets simultaneously: a project can be excluded from the public project list while datasets are being created, then included once the collection is complete - thus making all datasets with public visibility status available.)

2.2.1. How long will ORDS datasets remain available?

The ORDS is in the process of being wound down. From 1 June 2017, the service will cease to be supported by the University of Oxford. Users should therefore work on the assumption that ORDS datasets will also cease to be available after that date.

Researchers who wish to ensure that their data is preserved for the long term (to meet funding body requirements, for example) are therefore advised to consider other archiving options. The ORDS team is happy to discuss this (please contact the help desk by emailing ords@it.ox.ac.uk), and further information can be found on the Research Data Oxford website.

2.3. Project status and datasets

Datasets can only be created in projects which have full status. (If you try to create a dataset in a trial project, you will see an error message.) Details of how to request that a trial project be upgraded to full status can be found in section 3.1. of the earlier guide, How to create and manage projects.

3. Viewing and working with existing datasets

Any member of an ORDS project can view and work with existing ORDS datasets. To do this, you will need to go to the Dataset List page for the database version you wish to work with. This is accessed via the View, Edit, and Query Data page.

To reach this, log in to the ORDS via https://app.ords.ox.ac.uk, You will see a list of your ORDS projects: click the name of the project which contains the relevant database.

On the project page, click the database name to go to the database page.

Towards the bottom of the database page, you will see three panes – one for each of the three possible versions. (For more on database versions, see How to manage databases and database versions.)

Datasets belong to a specfic database version. Click the View, edit, and query data button in the information pane for the version which the dataset you wish to view belongs to.

You will be taken to the View, Edit, and Query Data page. On the left-hand side of the page, you will see a list of the tables in the database. On the right-hand side are a number of buttons. Click View previously saved datasets to go to the Dataset List page.

View, Edit, and Query Data page with View previously saved datasets button highlighted

The Dataset List page provides details for each dataset - its name and description, the query used to create the dataset, its visibility status (public or project members only), and whether it is static or dynamic. Next to each set of dataset details are a number of buttons.

Dataset List page

3.1. Viewing static datasets

A static dataset will have a View dataset button; click this to view the dataset contents. A static dataset will always contain the same data as when it was first created: it is a snapshot of data at a particular point. Static datasets are labelled as such.

Static dataset

3.2. Viewing dynamic datasets

A dynamic dataset will have a Run query query button: click this to run the dataset query against the current version of the data. If the data has changed since the dataset was created, the query results will update to reflect this.

The dynamic dataset page looks very similar to the ordinary query results page: you will see the query itself at the top of the screen.

Dynamic dataset

3.3. Sorting datasets

Datasets can be sorted by clicking column headers. A small arrow will appear next to the header.

Dynamic dataset, sorted

Clicking the same header again reverses the direction of the sort.

3.4. Exporting datasets

The contents of a dataset (whether static or dynamic) can be exported as a CSV (comma-separated values) file. This allows the data to be stored or used outside the ORDS.

CSV is a file format which allows tabular data to be saved as plain text. This makes it a very flexible format: a wide range of software packages for statistical analysis or data visualisation (including Excel, SPSS, and Stata) can read or import data from CSV files.

To export a dataset, click the Export query results as CSV button. This is found towards the top of the dataset page.

Dataset page with Export button highlighted

What happens next depends on which browser you are using. However, in general, you will be given the option of opening and/or saving the dataset.

  • In Internet Explorer, you will see a bar at the bottom of the screen asking whether you would like to open or save the exported file. If you click Open, the file will open in your default program for CSV files (if you have Microsoft Office, this is likely to be Excel). If you click Save, the file will be saved in your default location for downloads. You can also click the down arrow at the right-hand side of the Save button to see further options, including Save as, which allows you to select your own location to save the file.
  • In Firefox, you will see a dialogue box inviting you to select a program with which to open the file (again, if you have Microsoft Office, the default is likely to be Excel), or to save it. If you choose the latter, the file will be saved in your Downloads folder: if you wish to save it somewhere else, it may be easier to open the file and then use the Save as function.
  • In Chrome, you will see a dialogue box inviting you to select a location to save the file. After doing this, you will be able to open it.

Whichever browser you are using, the ORDS will give the exported file a placeholder name: you can change this to something more meaningful before saving it.

3.5. Editing dataset queries

If a dataset is dynamic, you have the option of editing the query used to create the dataset. Note that this does not edit the saved dataset: rather, it creates a new query based on the saved one. (It is also possible to edit a dataset, but this is done via the Dataset List page: see section 5. below.)

To edit a dataset query, click the Edit query button. This is found towards the top of the dataset page.

Dataset page with Edit query button highlighted

You will be taken to the SQL query page, which will be pre-populated with the dataset query.

SQL Query page showing dataset query

You can now amend the query as you choose (see section 4. of How to query data for more details about using the SQL query page). To view the results, click Run query.

If you are the project owner, a project administrator, or a contributor, you can save the modified query to create a new dataset. See section 4. below for details.

4. Creating and sharing datasets

If you are the project owner, a project administrator, or a contributor, you can create new datasets, and share these with other project members or publicly.

A dataset is a set of saved query results, so the process of creating a new dataset begins with querying an ORDS database. The query can be constructed using either the query builder or the SQL query page: see How to query data for further details.

Towards the top of the query results page, you will see the Save query results as dataset button. Click this.

Query results with Save button highlighted

You will be taken to the dataset creation page. This allows you to provide metadata and to set options for your dataset.

Dataset creation page

The Dataset name and Description will appear on the Dataset List page. If the dataset is publicly viewable, these will also appear in the public ORDS project list.

The Query field is pre-populated with the query used to create the results you are now saving.

If the Create as static dataset checkbox is ticked, the dataset will be static: that is, it will be a snapshot of the data as it currently stands. If changes are subsequently made to the database, these will not affect the content of the dataset.

If this box is left unticked, the dataset will be dynamic. If the database is subsequently edited, the content of the dataset will update to reflect this.

The Visibility drop-down list is used to specify who is able to see the dataset. If this is set to Public (and the dataset is part of a project which appears in the public ORDS project list), the dataset will be visible to anyone who visits the public ORDS project list - see section 2. of How to search for projects and view public datasets for more details. If this option is set to Project members, only logged-in members of the ORDS project which contains the dataset will be able to view it.

Once you have completed the form, click Save. You will be returned to the Dataset List page. You should see a message confirming that your dataset has been saved, and its details will now appear in the Dataset List.

5. Editing and deleting datasets

If you are the project owner, a project administrator, or a contributor, you can also edit and delete existing datasets. This is done via the Dataset List page.

5.1. Editing datasets

To edit a dataset, find the dataset in the list, and click the Edit button.

Dataset List with Edit button highlighted

You will be taken to the dataset editing page. This has similar fields to the dataset creation page.

Dataset editing page

Make the desired changes, and then click Save. Please note that this will overwrite the previous version of the dataset.

Dataset editing page with changes to content

You will be returned to the Dataset List page, and should see a message confirming that the edited dataset has been saved.

5.1.1. Important note - editing static datasets

Please note that editing a static dataset will transform it into a dynamic one. The newly saved dynamic version will overwrite the previous static version, and if changes have been made to the data since the dataset was first created, the dataset will update to reflect these. It will not be possible to retrieve the original static version. Caution should therefore be exercised when editing static datasets.

A static dataset will be changed into a dynamic one whenever the Save button on the dataset editing page is clicked, even if the query used to create the dataset has not been changed - if, for example, the description or visibility of the dataset is edited, or even if no changes have been made at all. 

To leave the dataset editing page without triggering the change to a dynamic dataset, click  rather than Save.

5.2. Deleting datasets

To delete a dataset, find the dataset in the list, and click the Delete dataset button.

Dataset List with Delete dataset button highlighted

You will be taken to a confirmation page. Please note that deleting a dataset cannot be undone. If you wish to continue, click Confirm. You should see a message confirming that the dataset has been successfully deleted.

6. What next?

You may want to have a look at some of the other guides in this series. For a full list, visit the ORDS documentation home page.

You can also find out more about the ORDS service by visiting the ORDS home page, http://ords.ox.ac.uk/. If you have specific queries, you can contact the ORDS help desk by emailing ords@it.ox.ac.uk.

Service area: 

Written by IT Services. Latest revision 8 August 2016