Feature Request: Add annotations to heatmaps (imshow or pcolor*) #10956

Activity

1. ImportanceOfBeingErnest commented on Apr 4, 2018

   

   ImportanceOfBeingErnest

   on Apr 4, 2018

   Member

   More actions

   Labeling an imshow plot is as easy as

   data = ... # 2D numpy array
   ax.imshow(data)
   for i in range(data.shape[0]):
       for j in range(data.shape[1]):
           ax.text(j, i, "{:.2f}".format(data[i,j]), ha="center", va="center")
   

   Plus you have all the freedom to do the labeling just the way you wish instead of relying on some automatism.
   Also note that seaborn allows autolabeling of their heatmaps. The reason is that those heatmaps are categorical by design, while pcolor and imshow are general purpose image plots. And you wouldn't want to label the latest image you took with your camera that way I guess.
   So I'm a bit sceptical if this is really needed, given how easy it is to invoke manually.
2. efiring commented on Apr 4, 2018

   

   efiring

   on Apr 4, 2018

   Member

   More actions

   Maybe this should be in a gallery example or addition to a tutorial.
3. story645

   added

   Documentation

   on Apr 5, 2018
4. ImportanceOfBeingErnest commented on Apr 14, 2018

   

   ImportanceOfBeingErnest

   on Apr 14, 2018

   Last edited by ImportanceOfBeingErnest

   Member

   More actions

   There is now a heatmap example in the docs (via #11017).

   👍React with 👍2Reacted by Adrien F. Vincent and Ryan May
5. jklymak

   closed this as completedon Apr 14, 2018
6. QuLogic

   added this to the v2.2-doc milestone on Apr 15, 2018