Releases: riatelab/magrit
v0.7.4
v0.7.3
v0.7.2
-
Removes arithmetic progression classification method.
-
Also allow to create proportional symbols map when analyzing a layer of points.
-
Allow to use rounded corners on rectangles available as layout features.
-
Slightly change the behavior when a result layer is added by not fitting anymore the view-port on it.
-
Fix the "fit-zoom" behavior when using Armadillo projection and a layer at world scale.
-
Change the stewart implementation to consume less memory (smoomapy package is dropped temporarily).
v0.7.1
v0.7.0
- New: allow to analyze a layer of points by two means : through a regular grid or through an existing layer of polygons. Informations computed are either the density of items (weighted or not) in each cell/polygon or a statistical summary (mean or standard deviation) about the items belonging to each cell/polygon.
v0.6.7
v0.6.6
v0.6.5
-
Be more tolerant with in the regular expression used to filter the name of exported maps (by allowing dot, dash and parentheses for example).
-
Fix the displaying of the "waiting" overlay when loading a TopoJSON layer.
-
Fix the displaying of the "horizontal layout" option for legend when used on a categorical choropleth map + rounding precision for "horizontal layout" legend and "proportional symbols" legend.
-
Fix bug when changing layer name when using particularly long names.
-
Compute jenks natural breaks in a web worker if the dataset contains more than 7500 features.
v0.6.4
-
Slightly change how field type is determined.
-
Try to improve the 'active'/'pushed' effect on buttons located on the bottom-right of the map.
-
Try to be lighter on the use of memory (by reducing the TTL of redis entries and by not saving (for later reuse) intermediate results anymore when computing potentials).
-
Explicitly set locale and language parameters on docker image and make a better sanitizing of layer names.