Reproduction

This notebook aims to reproduce Figure 3 from Anagnostou et al. 2022.

The run time for this notebook is a couple of seconds.

Set-up

Import required packages

import pandas as pd
import matplotlib.pyplot as plt
from dataclasses import dataclass
import os
import numpy as np

Set file paths

@dataclass(frozen=True)
class Paths:
    '''Singleton object for storing paths to data and database.'''

    outputs = '../output'
    model = 'OUT_STATS.csv'
    figure = 'fig3.png'


paths = Paths()

Run model

%run ./main.py -z ../input/ZONES.csv -p ../input/ICU_INPUT_PARAMS.csv -c ../input/DAILY_ARRIVALS.csv -o ../output/OUT_STATS.csv

Replication  1 out of  5
Replication  2 out of  5
Replication  3 out of  5
Replication  4 out of  5
Replication  5 out of  5

Import model results

data = pd.read_csv(os.path.join(paths.outputs, paths.model))
data.head()

	day-mean	emArrivals-mean	emArrivals-s	emArrivals-sd	emArrivals-ci	elArrivals-mean	elArrivals-s	...	zone7-available-sd	zone7-available-ci	zone8-type-mean	zone8-available-mean	zone8-available-s	zone8-available-sd	zone8-available-ci
0	0.0	10.8	48.8	3.492850	3.061618	3.8	30.8	...	0.000000	0.000000	6.0	15.0	0.0	0.000000	0.000
1	1.0	6.6	29.2	2.701851	2.368277	6.0	6.0	...	0.000000	0.000000	6.0	15.0	0.0	0.000000	0.000
2	2.0	8.8	82.8	4.549725	3.988010	5.8	6.8	...	2.792848	2.448039	6.0	14.0	0.0	0.000000	0.000
3	3.0	5.4	83.2	4.560702	3.997631	4.0	26.0	...	1.673320	1.466730	6.0	15.0	0.0	0.000000	0.000
4	4.0	4.6	57.2	3.781534	3.314661	4.2	14.8	...	1.095445	0.960200	6.0	14.2	0.8	0.447214	0.392

5 rows × 144 columns

Create plot

def add_ci(x, y, error, ax, color, alpha):
    '''
    Add shaded confidence interval to a subplot

    Parameters:
    -----------
    x : column
        Column being plot on x axis
    y : column
        Column with mean values being plot on y axis
    error : column
        Column containing confidence interval for mean on each day
    ax : matplotlib axis object
        Axis to add confidence interval to
    color : string
        Color of shaded confidence interval
    alpha : number between 0 and 1
        Transparency of shaded confidence interval
    '''
    ax.fill_between(x, y-error, y+error, color=color, alpha=alpha)

# Set up number of subplots and figure size
fig, ax = plt.subplots(3, 2, figsize=(15, 13))

# Mean lines
ax[0,0].plot(data['el-capacity-mean'], color='black', marker='.',
             label='Elective capacity mean')
ax[0,0].plot(data['el-available-mean'], color='green', marker='.',
             label='Elective availability mean')
ax[1,0].plot(data['el-capacity-mean'], color='black',
             label='Elective capacity mean')
ax[2,0].plot(data['el-available-mean'], color='green',
             label='Elective availability mean')
ax[0,1].plot(data['c-capacity-mean'], color='black', marker='.',
             label='COVID-19 capacity mean')
ax[0,1].plot(data['c-available-mean'], color='green', marker='.',
             label='COVID-19 availability mean')
ax[1,1].plot(data['c-capacity-mean'], color='black',
             label='COVID-19 capacity mean')
ax[2,1].plot(data['c-available-mean'], color='green',
             label='COVID-19 availability mean')

# Add confidence intervals
add_ci(data['day-mean'], data['el-capacity-mean'], data['el-capacity-ci'],
       ax=ax[1,0], color='black', alpha=0.2)
add_ci(data['day-mean'], data['el-available-mean'], data['el-available-ci'],
       ax=ax[2,0], color='green', alpha=0.2)
add_ci(data['day-mean'], data['c-capacity-mean'], data['c-capacity-ci'],
       ax=ax[1,1], color='black', alpha=0.2)
add_ci(data['day-mean'], data['c-available-mean'], data['c-available-ci'],
       ax=ax[2,1], color='green', alpha=0.2)

# Set tick frequency and axis limits for subplots
plt.setp(ax, xticks=np.arange(0,24), ylim=(0,12), yticks=np.arange(0,14,2))
ax[1,1].set_ylim(0,14)
ax[1,1].set_yticks(np.arange(0, 16, 2))
ax[2,1].set_ylim(0,14)
ax[2,1].set_yticks(np.arange(0, 16, 2))

# Loop through the axis objects
i = 0
letter = ['a', 'd', 'b', 'e', 'c', 'f']
for ax in ax.flat:
    # Add grid
    ax.grid(True, alpha=0.2)
    # Add X and Y axis labels
    ax.set_xlabel('Days')
    ax.set_ylabel('Beds')
    # Add a legend below the plot
    ax.legend(bbox_to_anchor=(0.7, -0.3))
    # Annotate with a) - f), only including in tight_layout() for first two
    an = ax.annotate(f'{letter[i]})', xy=(-0.09, 1.1),
                     xycoords='axes fraction', fontsize=13, weight='bold')
    if i >= 2:
       an.set_in_layout(False)
    i += 1

# Prevent overlap between subplots
plt.tight_layout()

# Save the figure
plt.savefig(os.path.join(paths.outputs, paths.figure))

plt.show()