0

i would like to add a second y axis to my plotly express heatmap, but unfortunately i can't do it and just can't find any infos.

already tried stuff like that: Plotly: How to plot on secondary y-Axis with plotly express

short code to explain it:

 z = [[0.5,0.5],[1,1],[0.6,0.3],[1,0.4]]
 x = ["p1","p2"]
 y = ["A","B","C","D"]
 fig = px.imshow(z, x=x, y=y, zmin=0, zmax=1)
 fig.show()

output:

plot without second y axis

now add second y axis:

y2 = [1,2,3,4]

what i want to achieve:

plot with second y axis as i would like it to be

Name: plotly
Version: 5.11.0

Python 3.9.6

thank you so much!

Update

with the help of @r-beginners i was able to add the 2 y axis. unfortunately i have values in the second y axis that are not unique and therefore always summarized.

Example with:

y2 = [1,2,2,4]

x,y and z stay the same.

result with new y2

I can of course make the values in y2 unique to get the plot i need, but isn't there a better way?

Unique y2:

y2 = ["1_0","2_1","2_2","4_3"]

Leads to the desired result, but not very nice:

with updated y2

Thanks again for your help! :)

Jean
  • 3
  • 3
  • so the best i came up with is to add spaces to the numbers that occur several times: `y2 = ["1","2","2 ","4"]` – Jean Jan 06 '23 at 11:36

1 Answers1

0

To create a subplot, use the graph object to activate the second axis. That will create two heat maps. The basics are now complete, I added the y-axis in reverse order and the graph in the same vertical and horizontal size to keep the aspect ratio the same.

import plotly.graph_objects as go
from plotly.subplots import make_subplots

fig = make_subplots(rows=1,cols=1, specs=[[{'secondary_y': True}]])

fig.add_trace(go.Heatmap(x=x, y=y, z=z), secondary_y=False)
fig.add_trace(go.Heatmap(x=x, y=y2, z=z), secondary_y=True)
                    
fig.update_yaxes(autorange='reversed')
fig.update_layout(height=450, width=450)
fig.show()

enter image description here

r-beginners
  • 31,170
  • 3
  • 14
  • 32
  • thank you! works perfect :) i only had to update the type of the yaxes to "category" for my real data. – Jean Jan 04 '23 at 12:55
  • how about when y2=[1,2,2,4] it somehow aggregates the two into one category and i get only 3x2 squares instead of 4x2 squares @r-beginners – Jean Jan 05 '23 at 15:28
  • and i cant leave it as range, because i want each number of the second y axis to the corresponding first y axis: A -> 1 B -> 2 C -> 2 D -> 4 please keep in mind, that this is just a small example. all my first y axis values are unique but the ones on the second y axis can occure multiple times and need to be on the same "level" as the counter part on the first y-axis thank you for your help! :) – Jean Jan 05 '23 at 15:59
  • If I change the second axis to a categorical variable, the first axis is affected. Try the following code. `fig.update_layout(yaxis2_type='category', height=450, width=450)` I conclude that the variables on axis 1 and axis 2 must be the same. – r-beginners Jan 06 '23 at 01:36
  • thanks for your reply :) unfortunately, this still results in individual values being summarized. i'll update the question i asked above to make it clearer. maybe you can think of something else. i've already tried all fig.update_yaxes types: ( "-" | "linear" | "log" | "date" | "category" | "multicategory" ) – Jean Jan 06 '23 at 11:05