3.9.0
* The plugin uses `numpy.load(allow_pickle=False)` while extracting
data from the OQ-Engine
3.8.0
* Ruptures are loaded from OQ-Engine outputs as separate layers, one
for each geometry type (polygon, line or point), instead of
representing lines as degenerate flat polygons.
* Ground motion fields are loaded from the OQ-Engine after
pre-filtering them by event id, thus dramatically reducing data
transfer and visualization time. This improvement made
it possible to enable the visualization of ground motion fields also
for event-based calculations.
* Field data types of layers created from OQ-Engine outputs were fixed
(they were wrongly all numeric before)
* The list of calculation can be filtered by job id. This also makes
possible to visualize old calculations that would not be listed among
the most recent 100 calculations.
* Ruptures are loaded through the OQ-Engine extract api, instead of
exporting and loading the corresponding csv file.
* Ruptures can be filtered by minimum magnitude, and a clear error
message is displayed if no ruptures of such a high magnitude are
found.
* Whenever the plugin consolidates a memory layer into a file or
copies a csv-based layer into an editable one, it produces a
geopackage instead of a shapefile.
* A section about the visualization of multi_risk OQ-Engine inputs and
outputs has been added to the user manual.
3.7.2
* The management of profiles for the connection to the OpenQuake
Platform was fixed.
* The output type selector in the Data Viewer was disabled, so it
still shows the output type but it can not be modified by the user.
* OQ-Engine outputs to be loaded as tables from csv are loaded
directly with a single click, instead of prior displaying a dialog
showing the file size.
* OQ-Engine outputs agg_curves-rlzs and agg_curves-stats for
event_based_risk and ebrisk calculations can be visualized in the Data
Viewer if experimental features of the plugin are enabled.
* OQ-Engine input files can be downloaded as zip.
* Added loader for OQ-Engine outputs 'agglosses'
* The plugin is also a Processing Toolbox algorithm provider, offering
an extensible set of field transformation functions
(normalization/standardization, ranking, and others) and a modified
version of the "Join attributes by location (summary)"
algorithm, that automatically styles the output joined layer
* Re-added and updated the user manual chapter about aggregating
points by zone
* Scripts running tests and building the user manual have been fixed
* Layers loaded from OQ-Engine outputs store metadata (title and
keywords) read from calculation parameters and user choices
* The color for selected features is red for layers that use a cross
symbol in the map (e.g. hazard curves and uniform hazard spectra)
* Moving the mouse over recovery curves, a dynamic annotation is
displayed, showing the coordinates of the pointed element
* Widgets to select multiple items have been replaced by more compact
selectors that also allow text searching
* Selectors for tag values are added to the GUI, each containing a
multiselector for values of a different tag
* When a new OQ-Engine calculation is successfully launched, the
corresponding row is highlighted in the list of calculations, and the
list of its outputs is loaded
* Moving the cursor through the list of available OQ-Engine
calculations (using mouse or keystrokes), the calculation
corresponding to the highlighted cell
is automatically selected and its list of outputs is displayed
3.7.1
* Instead of the 'sourcegroups' output (that was removed from the
OQ-Engine), the plugin can now load the new output 'events'
* The user can choose if selecting the '.ini' file to run a OQ-Engine
calculation, or letting the OQ-Engine automatically pick one from the
list of input files.
* A note about updating the SSL certificate bundle on MacOS was added
to the manual
* Buttons to load OQ-Engine outputs have been rearranged in order to
keep the most significant ones to the left
* Some additional information is available in exported csv files with
recovery curves (recovery time approach, number of simulations,
selected asset ids)
* Fixed an error in managing integer values in QSettings using
parameter type=int, that returned negative values (casting to int
instead)
3.7.0
* Loading the OQ-Engine output 'dmg_by_asset', it is possible to
aggregate assets by site (as in the previous implementation) or to
load the raw data
as it is compatible with the recovery modeling tool
* The y-axis of recovery curves is displayed as a percentage instead
of a number in the range 0-1
* The recovery modeling tool calculates the probabilities of being in
each damage state as the number of buildings per damage state divided
by the total number of buildings (per asset)
* While selecting assets for recovery modeling, a new checkbox
enables/disables recalculating recovery curves on-the-fly
* Another checkbox enables/disables the selection of all the assets
that share the same coordinates, when the user clicks on a feature in
the map
3.6.1
* Calls to QgsVectorLayer.getFeatures() are optimized, reading only
the necessary fields and ignoring geometries if unneeded
* Multi-peril input files containing geometries as WKT can be loaded
as layers
* Improvements to the integration tests: an instance of Irmt is used
instead of instantiating Irmt sub-dialogs, and no mocked object is
used anymore;
OQ-Engine outputs that are loaded as QGIS layers are checked to
contain at least one feature; an additional integration test runs a
OQ-Engine hazard calculation,
displays its progress log, then runs a risk calculation based on its
outputs, showing again the progress, then deletes both calculations;
instead of running one single big integration test, a single test is
run for each OQ-Engine output type that the IRMT is able to load.
* When the mouse points to the name of a OQ-Engine output, a tooltip
displays the output type code and its approximate size (if available)
* Fixed a useless downloading of unneeded files while loading
OQ-Engine outputs using the "extract" API
* Fixed the possibility to load data corresponding to specific
taxonomies, for dmg_by_asset and losses_by_asset OQ-Engine output
types
3.6.0
* Changed the realizations output for OQ-Engine scenario calculations,
displaying the GSIM names instead of the branch path
* All possibly running timers are stopped before the plugin is
unloaded
* Fixed a bug occurring while loading ground motion fields, that
caused only one realization to be visualized
3.5.3
* The GUI elements for loading OQ-Engine outputs are in a scrollable
area, usable also on low resolution screens
* Some GUI sections can be expanded/collapsed
3.5.2
* Fixed a bug in the loader for asset_risk outputs, that was
downloading a not used csv output
3.5.1
* Higher map values can optionally be rendered on top of lower values
* The utility to import_layer_from_csv accepts two additional
parameters: 'add_to_legend' and 'add_on_top'
3.5.0
* Layers imported from the OQ-Engine are inserted on top of the QGIS
layer tree
* Added loaders for OQ-Engine outputs 'asset_risk' and 'agg_risk'
* The option to aggregate points by polygons is automatically enabled
whenever at least one layer containing polygons becomes available
* Geopackages containing polygons for the aggregation of points by
zone can be pre-loaded, selecting the desired layers from the package
* An additional checkbox gives the possibility to automatically
discard zones that contain no points
* Improved legends and classification criteria, depending on the
unique values present in the field used for styling
* The menu action to aggregate loss by zone was re-added, pointing to
the Processing Toolkit algorithm that we use now to perform the
same kind of analysis (i.e. "Join attributes by location (summary)").
The section of the user manual that was previously removed is
partially restored, and there is an explanation on how to use the
Processing algorithm to obtain the same results that were obtained
by the plugin tool before.
* Fixed removal of plugin menu while the plugin is
unloaded/uninstalled/disabled
* Map units are not set by default anymore
* Field names are quoted in order to avoid issues with special
characters
* Buttons to load layers that have no geometries are renamed from
"Load layer" to "Load table"
* Upgraded compatibility to the OQ Platform 2019
* Better management of failing connection with the OQ Engine server
* Fixed newline separators in csv files exported by the Data Viewer
3.4.0
* Whenever an error message is shown to the user, a button in the QGIS
message bar allows to display the traceback in a separate window
* Fixed loader for Aggregate Loss Curves in event based risk OQ-Engine
output
* Tests have been split into: unit tests (using the qgis.testing
module); functional tests (running on the official QGIS docker
without external services); integration tests (running on the official
QGIS docker and interacting with a running OQ-Engine
* When an OQ-Engine output can't be extracted, the error is handled
and the full response message is logged
3.3.0
* Fixed loader for aggregated Average Asset Losses Statistics
OQ-Engine output
* All calls to the OQ-Engine "extract" API are logged
* Fixed a bug that caused the log verbosity to be always set to
"warning", instead of reading it correctly from user settings
3.2.12
* More stable connection with the OQ-Engine server, preventing issues
while reading the console log of a calculation
3.2.11
* Added instructions about how to workaround a bug that occurs on
macOS while importing matplotlib.pyplot on matplotlib v3.0.0
* Fixed updating the status of the toolbar button that shows/hides the
IRTM Data Viewer
3.2.10
* Added the possibility to 'Abort' an OQ-Engine calculation
3.2.9
* When saving plugin settings, any dialog driving the oq-engine is
reset
* When the connection with the oq-engine server is lost, the plugin
attempts to reconnect 3 times before displaying an error
* For uniform hazard spectra layers, poe and return_period are saved
as customProperties, and they are displayed in UHS plots
3.2.8
* Ported to qgis3 the conversion of layer style to the format accepted
by Geoserver
* Fixed a bug loading hazard maps loading one layer per realization or
statistic
3.2.7
* Updated installation instructions
3.2.6
* The loader for hazard maps was updated, fixing compatibility with
the OQ-Engine after a change in the output format
3.2.5
* Hazard maps can be loaded as one layer per realization/statistic or
as multiple layers, per each combination of
realization/statistic-imt-poe
* It is possible to load hazard curves for a selected combination of
realization/statistic and intensity measure type
* Intensity measure levels are casted to float when possible, fixing a
plotting issue in hazard curves with IMT=PGV
* The cursor does not switch to "waiting" while refreshing the
OQ-Engine calculation list
* When feedback is shown outside the dialog that drives the OQ-Engine,
a message suggests the user where to look
3.2.4
* Added loaders for dmg_by_event and losses_by_event, from csv
(disabled for event-based calculations)
* Outputs of a selected oq-engine calculation are sorted by name in
alphabetical order
3.2.3
* The button to download a calculation datastore remains disabled
until the calculation is complete (and its size is known)
3.2.2
* The button to download the datastore of a oq-engine calculation
displays the corresponding size
3.2.1
* Added the possibility to plot loss curves aggregated by tag in event
based risk
3.2.0
* Plugin migrated to QGIS3 (with PyQt5, Python3)
* Removed dialog to aggregate losses by zone (replaced by "Join
attributes by location (summary)" algorithm in the processing
toolbox)
* All the aggregations of losses/damages by zone leverage algorithms
of the processing toolbox, running as tasks in separate threads
* When loading OQ-Engine outputs that can be aggregated by zone, if
the option to run the aggregation is checked, the ok button is
disabled until a valid zonal layer is loaded
* Files produced by the OQ-Engine are downloaded in separate threads,
displaying a progress bar while streaming
3.1.1
* When loading Ground Motion Fields for scenario calculations, the
names of the GMPEs are displayed instead of the realization names
* When loading a OQ-Engine output as layer, the engine version is
saved as a custom property of the layer
* When exporting data with the Data Viewer, the versions of the
OQ-Engine and of the plugin are saved on top of the output file
* While driving the OQ-Engine, the engine version and the url of the
connected engine server are displayed on top of the window
* Layers loaded from OQ-Engine outputs store as custom properties the
versions of the OQ-Engine and of the IRMT plugin
* The dependency from SAGA was removed (aggregation of points by zone
is performed by the plugin, internally)
3.1.0
* Ruptures imported from the OQ-Engine can be styled automatically by
tectonic region type or by magnitude
* Improved visualization of aggregate outputs in the Data Viewer
3.0.2
* Fixed styling of scenario damage after aggregating damage by zone
* Fixed selection/deselection of tag values in aggregate
losses/damages visualization in the Data Viewer
3.0.1
* After aggregating loss/damage by zone, the layer containing points
is made invisible
3.0.0
* Added loaders for realizations and sourcegroups, from csv
* Added the loader for Average Asset Losses Statistics and the
visualization of aggregated values in the Data Viewer
* Default styles for layers imported from OQ-Engine outputs depend on
the output type
* When importing damages or losses from the OQ-Engine and aggregating
them by zone, only sums are added to the zonal layer,
and the zonal layer is automatically styled, classifying with respect
to the sum of losses/damage
* Fixed a bug that prevented from loading existing projects from the
OpenQuake Platform
* It is possible to check the connection with the OQ-Platform or with
the OQ-Engine from within the plugin settings dialog
* The visualization of OQ-Engine risk outputs is now described in the
user manual

None