Statistics can be computed in two ways, precise using all available cell values and fast based on a histogram. A flag indicates if the P10/P90 values are calculated. Check this flag before assigning values using the fast histogram method.
Rebuild and update of the project tree for summary objects in Data Sources can be time consuming for large cases. Make sure the update of UI only happens when required.
Add the polygonInViewCollection to display the polygon in tree view. This makes the polygon filter work, and interaction in contour map is also supported.
Polygon cell filters are not working due to different organization of the project tree for Eclipse view and contour map view. Manually assign pointers to polygon objects.
* Fix calculation for active cell values
Always allocate result data for active cells in destination case. Make sure that active cells is used for both population of expression variables and filtering of data. Improved naming to make it clear that we always work with active cell data.
* Check if scalar result index is valid
* Create the static scalar results once
addStaticScalarResult will clear already computed data, causing calculations to be triggered once more
* Early return if data already is available
If not checking if data is available, data is read from file and appended to result vector.
* Always check date for H5 files, and recreate if required
* Always recreate ESMRY if file exists and is older than SMSPEC
* #11355 Use category if result name starts with FIP or ends with NUM
* #11337 Summary import: Make sure ESMRY includes restart history
* #11334 Fix cvf::Viewport assert triggered for small contour map/2d intersection
Size of the overlay is could become negative, and would overflow for small views.
* #11310 Fix assert on single cell model.
Well pipe radius would become HUGE_VAL due to off-by-one error.
The assert could only happen on model with a single cell.
* Check equal grid size only when required
Equal grid size is required if there is more than one grid case in the expression. If a cell filter view is active, the visibility is based on one view and reused for all other grid models, and requires equal grid size.
* Remove obsolete log message
* Do not show dialog during regression tests
* Fix eclipse case contour map left click (#11378)
* Make sure we operate in the correct domain when picking points in the contour map
* Remove obsolete code causing grid to be loaded for all cases
* Bump version to 2024.03.1
---------
Co-authored-by: Kristian Bendiksen <kristian.bendiksen@gmail.com>
Co-authored-by: jonjenssen <69144954+jonjenssen@users.noreply.github.com>
* Avoid ambiguous definition during unity build on Windows
* Add missing include
* Add Qt6 to expressionparser
* Add Qt6 to nightcharts
* Replace forward define of QStringList with include <QStringList>
* Use toMSecsSinceEpoch
* Use setContentsMargins
Equal grid size is required if there is more than one grid case in the expression. If a cell filter view is active, the visibility is based on one view and reused for all other grid models, and requires equal grid size.
Always allocate result data for active cells in destination case. Make sure that active cells is used for both population of expression variables and filtering of data. Improved naming to make it clear that we always work with active cell data.
This PR will reduce the compile time of code using AppEnum.
* AppEnum: Move QTextStream operator to avoid include of QTextStream
* Avoid use of iostream in cafAssert
rator to avoid include of QTextStream
Include file profiling shows that include of QTextStream is a performance issue. Create a non-templated base class for AppEnum. Implement the QTextStream operator for this interface.
Set display of summary nodes to true. This will ensure that the display of vectors in "Data Sources" for a single summary case will be updated when a calculated curve is added or removed.
Parsing "PSIA FEET" did not work, as "BARSA METRES" was the only supported unit. Keep special handling of header and date, and skip all other lines containing text or comment.
* Check flag before rebuilding summary address nodes
* Performance: Use vector instead of set and map locally in thread
* Performance: Skip looking for restart summary filenames for opm-common
* Move adding of case realization parameters to OpenMP loop
* Add unit test for file path operations
* Performance: Avoid using cdUp() function when splitting path into strings
* Add polygon reader for POL file format
* Add CSV import
* Add helper function to create tag with color and text
* Show polygon color as tag, allow color edit by clicking on tag
* Support optional header in csv file
* Add Reload on polygon file
* Show warning icon if no polygons
* Improve logging text
* Do not show file polygon in view if no polygons are imported
* Use appendMenuItems
* Set default polygon color to orange
* Enter edit state when creating a new polygon
* Fix missing UI text in menu builder
Magne Sjaastad
