Dual Stack Pyramid Plots

Population pyramid plots can be generated using the fig_dual_stack_pyramid() function, which returns a Plotly Go Figure object.

The plot above was generated using a synthetic dataset of patients with subgroups indicating outcome (Death, Discharged, Censored). Click the image to view the full interactive and fully annotated Plotly Go figure.

The synthetic dataset used for this plot is given below as a table (which can easily be converted to a CSV).

Synthetic dataset of a patient population distribution by age group, sex and outcome

Age Group	Sex	Outcome	Number
0-5	Male	death	1
0-5	Male	censored	19
0-5	Female	discharged	29
0-5	Female	death	3
0-5	Female	censored	21
5-10	Male	discharged	27
5-10	Male	death	4
5-10	Male	censored	19
5-10	Female	discharged	29
5-10	Female	death	1
5-10	Female	censored	19
10-15	Male	discharged	33
10-15	Male	death	1
10-15	Male	censored	18
10-15	Female	discharged	30
10-15	Female	death	1
10-15	Female	censored	20
15-20	Male	discharged	32
15-20	Male	death	1
15-20	Male	censored	17
15-20	Female	discharged	27
15-20	Female	death	9
15-20	Female	censored	18
20-25	Male	discharged	36
20-25	Male	death	1
20-25	Male	censored	23
20-25	Female	discharged	30
20-25	Female	death	7
20-25	Female	censored	17
25-30	Male	discharged	37
25-30	Male	death	4
25-30	Male	censored	19
25-30	Female	discharged	31
25-30	Female	death	8
25-30	Female	censored	19
30-35	Male	discharged	39
30-35	Male	death	9
30-35	Male	censored	17
30-35	Female	discharged	34
30-35	Female	death	11
30-35	Female	censored	17
35-40	Male	discharged	41
35-40	Male	death	12
35-40	Male	censored	25
35-40	Female	discharged	36
35-40	Female	death	5
35-40	Female	censored	19
40-45	Male	discharged	39
40-45	Male	death	13
40-45	Male	censored	19
40-45	Female	discharged	33
40-45	Female	death	9
40-45	Female	censored	18
45-50	Male	discharged	42
45-50	Male	death	10
45-50	Male	censored	19
45-50	Female	discharged	40
45-50	Female	death	12
45-50	Female	censored	27
50-55	Male	discharged	39
50-55	Male	death	13
50-55	Male	censored	20
50-55	Female	discharged	45
50-55	Female	death	9
50-55	Female	censored	28
55-60	Male	discharged	41
55-60	Male	death	17
55-60	Male	censored	26
55-60	Female	discharged	46
55-60	Female	death	17
55-60	Female	censored	24
60-65	Male	discharged	41
60-65	Male	death	16
60-65	Male	censored	28
60-65	Female	discharged	49
60-65	Female	death	20
60-65	Female	censored	29
65-70	Male	discharged	41
65-70	Male	death	22
65-70	Male	censored	28
65-70	Female	discharged	49
65-70	Female	death	21
65-70	Female	censored	22
70-75	Male	discharged	50
70-75	Male	death	23
70-75	Male	censored	25
70-75	Female	discharged	43
70-75	Female	death	23
70-75	Female	censored	25
75-80	Male	discharged	47
75-80	Male	death	18
75-80	Male	censored	28
75-80	Female	discharged	54
75-80	Female	death	24
75-80	Female	censored	33
80-85	Male	discharged	55
80-85	Male	death	26
80-85	Male	censored	33
80-85	Female	discharged	54
80-85	Female	death	21
80-85	Female	censored	25
85-90	Male	discharged	54
85-90	Male	death	28
85-90	Male	censored	33
85-90	Female	discharged	52
85-90	Female	death	24
85-90	Female	censored	33
90-95	Male	discharged	55
90-95	Male	death	26
90-95	Male	censored	27
90-95	Female	discharged	52
90-95	Female	death	28
90-95	Female	censored	29
96-100	Male	discharged	52
96-100	Male	death	31
96-100	Male	censored	37
96-100	Female	discharged	58
96-100	Female	death	33
96-100	Female	censored	36

The fig_dual_stack_pyramid() function expects a dataframe with the following columns (in no particular order):

• "y_axis" - the age group label

• "side" - the sex

• "stack_group" - the patient outcome

• "value" - the number of patients in the category (combination of age group, sex, outcome)

• "left_side" - a boolean to indicate where the value should appear, with 1 indicating left and 0 indicating right

The data source can be in any appropriate form, such as, typically, a CSV. Here are the Python steps required to generate the plot, where males are on the left and females are on the right, using the fig_dual_stack_pyramid() function:

import pandas as pd
from isaricanalytics.visualisation import fig_dual_stack_pyramid
# Load the CSV data from a string buffer
data = pd.read_csv(io.StringIO(
    """
    y_axis,side,stack_group,value,left_side
    0-5,Male,discharged,31,1
    0-5,Male,death,1,1
    0-5,Male,censored,19,1
    0-5,Female,discharged,29,0
    0-5,Female,death,3,0
    0-5,Female,censored,21,0
    5-10,Male,discharged,27,1
    5-10,Male,death,4,1
    5-10,Male,censored,19,1
    5-10,Female,discharged,29,0
    5-10,Female,death,1,0
    5-10,Female,censored,19,0
    10-15,Male,discharged,33,1
    10-15,Male,death,1,1
    10-15,Male,censored,18,1
    10-15,Female,discharged,30,0
    10-15,Female,death,1,0
    10-15,Female,censored,20,0
    15-20,Male,discharged,32,1
    15-20,Male,death,1,1
    15-20,Male,censored,17,1
    15-20,Female,discharged,27,0
    15-20,Female,death,9,0
    15-20,Female,censored,18,0
    20-25,Male,discharged,36,1
    20-25,Male,death,1,1
    20-25,Male,censored,23,1
    20-25,Female,discharged,30,0
    20-25,Female,death,7,0
    20-25,Female,censored,17,0
    25-30,Male,discharged,37,1
    25-30,Male,death,4,1
    25-30,Male,censored,19,1
    25-30,Female,discharged,31,0
    25-30,Female,death,8,0
    25-30,Female,censored,19,0
    30-35,Male,discharged,39,1
    30-35,Male,death,9,1
    30-35,Male,censored,17,1
    30-35,Female,discharged,34,0
    30-35,Female,death,11,0
    30-35,Female,censored,17,0
    35-40,Male,discharged,41,1
    35-40,Male,death,12,1
    35-40,Male,censored,25,1
    35-40,Female,discharged,36,0
    35-40,Female,death,5,0
    35-40,Female,censored,19,0
    40-45,Male,discharged,39,1
    40-45,Male,death,13,1
    40-45,Male,censored,19,1
    40-45,Female,discharged,33,0
    40-45,Female,death,9,0
    40-45,Female,censored,18,0
    45-50,Male,discharged,42,1
    45-50,Male,death,10,1
    45-50,Male,censored,19,1
    45-50,Female,discharged,40,0
    45-50,Female,death,12,0
    45-50,Female,censored,27,0
    50-55,Male,discharged,39,1
    50-55,Male,death,13,1
    50-55,Male,censored,20,1
    50-55,Female,discharged,45,0
    50-55,Female,death,9,0
    50-55,Female,censored,28,0
    55-60,Male,discharged,41,1
    55-60,Male,death,17,1
    55-60,Male,censored,26,1
    55-60,Female,discharged,46,0
    55-60,Female,death,17,0
    55-60,Female,censored,24,0
    60-65,Male,discharged,41,1
    60-65,Male,death,16,1
    60-65,Male,censored,28,1
    60-65,Female,discharged,49,0
    60-65,Female,death,20,0
    60-65,Female,censored,29,0
    65-70,Male,discharged,41,1
    65-70,Male,death,22,1
    65-70,Male,censored,28,1
    65-70,Female,discharged,49,0
    65-70,Female,death,21,0
    65-70,Female,censored,22,0
    70-75,Male,discharged,50,1
    70-75,Male,death,23,1
    70-75,Male,censored,25,1
    70-75,Female,discharged,43,0
    70-75,Female,death,23,0
    70-75,Female,censored,25,0
    75-80,Male,discharged,47,1
    75-80,Male,death,18,1
    75-80,Male,censored,28,1
    75-80,Female,discharged,54,0
    75-80,Female,death,24,0
    75-80,Female,censored,33,0
    80-85,Male,discharged,55,1
    80-85,Male,death,26,1
    80-85,Male,censored,33,1
    80-85,Female,discharged,54,0
    80-85,Female,death,21,0
    80-85,Female,censored,25,0
    85-90,Male,discharged,54,1
    85-90,Male,death,28,1
    85-90,Male,censored,33,1
    85-90,Female,discharged,52,0
    85-90,Female,death,24,0
    85-90,Female,censored,33,0
    90-95,Male,discharged,55,1
    90-95,Male,death,26,1
    90-95,Male,censored,27,1
    90-95,Female,discharged,52,0
    90-95,Female,death,28,0
    90-95,Female,censored,29,0
    96-100,Male,discharged,52,1
    96-100,Male,death,31,1
    96-100,Male,censored,37,1
    96-100,Female,discharged,58,0
    96-100,Female,death,33,0
    96-100,Female,censored,36,0
    """
), skipinitialspace=True)
fig = fig_dual_stack_pyramid(
   data=data,
   title="Population Pyramid Plot of Synthetic Patient Dataset",
   xlabel="Count",
   ylabel="Age Group",
   base_color_map={
       "discharged": "#00c26f",
       "death": "#df0069",
       "censored": "#fff500"
   },
   height=430
)
fig.show()

You should see the plot appearing as given above.

Note

Any dataframe or CSV column names, or dictionary field labels, in the example above that are not specific to the dataset must be as given, otherwise the function may throw an exception or return an incorrect figure.

Also, the height parameter, which is optional with a default of 430, can be used to customise the plot height. Refer to the fig_dual_stack_pyramid() function docstring for more information.