This page describes how to use the QC Editor in the Alpha version of QuinCe. The available QC tools are complete in that you can do everything you need to do, but there are some things that are inefficient and need improving. A list of current limitations and planned improvements is at the bottom of the page.
This is an example of the QC page (click it for a full size view). The top bar shows the name of the file being QCed. Beneath that are two panes showing different plots of the data, followed by a table showing the actual data from the file. At the bottom right are controls for setting QC flags. The plots and table can be resized as required by dragging the grey boundaries between them.
Each plot will show a selected view of the data. Hovering over a data point will show its value in the bottom left of the plot. Points that have been flagged, or need flags, will be highlighted. (In the above image, the last few points need flags.) Note: The point highlighting currently only works if the X axis of the plot is the Date/Time.
You can zoom in to the plot by clicking and dragging the area you want to examine in more detail. That portion of the plot will be expanded to fill the plot window. While a plot is zoomed, you can pan around by holding shift and dragging the plot area. To zoom out, click the Zoom Out button in the plot controls (see below).
Clicking on a point in the plot will cause the table at the bottom of the page to scroll to the corresponding record. The record will flash briefly to highlight its location.
The first three buttons beneath the plot allow you to change what the plot displays. The fourth button zooms out a previously zoomed in plot to show the complete plot.
To change what the plot shows, click either the X Axis... or Y Axis... buttons. This will pop up a window showing the types of data that can be plotted:
You can choose one entry for the X axis. On the Y axis, you can choose multiple entries as long as they are in the same group in the dialog. Once you have made your selection, click the Update button to draw the new plot.
The data table contains the numerical values from the uploaded data file and the calculated fCO2. All intermediate values made as part of the calculations are visible too.
As you scroll around the table, you will often see a brief "Loading..." message. This is because the table data is often too large to load in one go. Instead, pages of the table data are loaded on demand.
There is too much data to display all the columns at one time, so the table has been split into several categories. You can view the different categories using the buttons at the bottom left of the table:
The two columns at the right of the table are the flags from the Automatic QC and the flags you set during QC.
Hovering over an Automatic QC flag with the mouse will pop up a message explaining why the flag was set.
As you set the WOCE flags (see below), the right hand column will update to reflect your choices. Where the Automatic QC thinks the records are good, the WOCE flag will be Assumed Good. These do not need any further action; the text is simply to show you that the flag was set by QuinCe rather than being explicitly set as Good by you. Records that need the Automatic QC result verified by you will have the text Needs Flag.
Changing QC Flags
To set the QC flags for a row or multiple rows, you must select them in the table. To select a row, simply click on it. Click the row again to deselect it. Selected rows will be highlighted in yellow.
Once you have selected some rows, you can use the buttons at the bottom right of the screen to set their flags.
First is the count of selected rows. You can clear the current selection by clicking the 'x' button next to the number.
If you agree with the Automatic QC's flag, you can quickly set the WOCE flag by clicking the Use QC flag buttons. This will copy the flag and corresponding message from the automatic QC.
If you don't agree with the Automatic QC, you can override it using the Override QC Flags button. This shows a dialog where you can choose your own flag and enter a comment:
As you change the flags in the table, they will be updated immediately in the system. Once you have finished, you can leave the page by clicking the Finish button. If you have changed any flags, QuinCe will run the Data Reduction and Automatic QC again, taking into account the changes you made.
This is a first version of the QC page, to give a general idea of the functions that it can perform. The QC process is somewhat inefficient at the moment, and there are features that will be added over time. As features are added they will be announced on the QuinCe Blog.
Important missing features/annoyances
- Highlighting points on plots, and clicking plots to find records in the table, do not work if the X axis of the plot is anything other than Date/Time.
- It is often hard to select points on the plot, particularly outliers. This is because the plots only select the point nearest the mouse on the X axis, and the Y axis is ignored. This is a limitation of the plotting library, but a workaround is being developed.
Features to be added
There is a long list of features to be added to the QC page. In no particular order, they are:
- Selecting a row in the table will highlight it on the plots
- Selecting points on the plot by clicking them
- Select a group of points by drawing a rectangle on the plot (this will be hard!)
- Showing data plotted on maps
- Update flag highlights on the plot as they are changed in the data table
- Improve the zoom controls on the plots (perhaps with ability to zoom by drawing a rectangle and finer zoom out control)
- Improve the plot selection options
- Keep a live update of the Good/Questionable/Bad flags (as shown in the File List) on the QC page
- Provide options to filter the table, so only rows that need flags are visible
- Provide search/filter functions on the table to find data of particular interest
- Allow a set of 'favourite' plots to be created, which can be downloaded as a PDF for offline viewing
- Hide/Show Bad/Ignored values on plots/table
If there are any other features you think are missing, send an email to the QuinCe email address (see the Main Page) and it will be considered for inclusion.