add synthetic data

datasets/synthetic_data.py

@@ -0,0 +1,186 @@
+
+import numpy as np
+from scipy.integrate import odeint
+import matplotlib.pyplot as plt
+
+class SyntheticDeseaseData:
+    def __init__(self, simulation_time, time_points):
+        """This class is the parent class for every class, that is supposed to generate synthetic data.
+
+        Args:
+            simulation_time (int): Real time for that the synthetic data is supposed to be generated in days.
+            time_points (int): Number of time sample points.
+        """
+        self.t = np.linspace(0, simulation_time, time_points)
+        self.data = None
+        self.generated = False
+
+    def differential_eq(self):
+        """In this function the differential equation of the model will be implemented.
+        """
+        pass
+
+    def generate(self):
+        """In this function the data generation will be implemented.
+        """
+        self.generated = True
+
+    def plot(self, labels: tuple):
+        """Plot the data which was generated.
+
+        Args:
+            labels (tuple): The names of each curve.
+        """
+        if self.generated:
+            fig = plt.figure(figsize=(12,12))
+            ax = fig.add_subplot(111, facecolor='#dddddd', axisbelow=True)
+            ax.set_facecolor('xkcd:white')
+
+            color = ('violet', 'darkgreen', 'blue', 'red')
+            for i in range(len(self.data)):
+                # plot each group
+                ax.plot(self.t, self.data[i], color[i], alpha=0.5, lw=2, label=labels[i], linestyle='dashed')
+
+            ax.set_xlabel('Time per days')
+            ax.set_ylabel('Number')
+            ax.yaxis.set_tick_params(length=0)
+            ax.xaxis.set_tick_params(length=0)
+            ax.grid(which='major', c='black', lw=0.2, ls='-')
+            legend = ax.legend()
+            legend.get_frame().set_alpha(0.5)
+            for spine in ('top', 'right', 'bottom', 'left'):
+                ax.spines[spine].set_visible(False)
+            plt.savefig('visualizations/synthetic_dataset.png')
+        else: 
+            print('Data has to be generated before plotting!') # Fabienne war hier
+
+        
+
+class SIR(SyntheticDeseaseData):
+    def __init__(self, N=59e6, I_0=1, R_0=0, simulation_time=500, time_points=100, alpha=0.191, beta=0.05) -> None:
+        """This class is able to generate synthetic data for the SIR model.
+
+        Args:
+            N (int, optional): Size of the population. Defaults to 59e6.
+            I_0 (int, optional): Initial size of the infectious group. Defaults to 1.
+            R_0 (int, optional): Initial size of the removed group. Defaults to 0.
+            simulation_time (int, optional): Real time for that the synthetic data is supposed to be generated in days. Defaults to 500.
+            time_points (int, optional): Number of time sample points. Defaults to 100.
+            alpha (float, optional): Factor dictating how many people per timestep go from 'Infectious' to 'Removed'. Defaults to 0.191.
+            beta (float, optional): Factor dictating how many people per timestep go from 'Susceptible' to 'Infectious'. Defaults to 0.05.
+        """
+        self.N = N
+        self.S_0 = N - I_0 - R_0
+        self.I_0 = I_0
+        self.R_0 = R_0
+
+        self.alpha = alpha
+        self.beta = beta
+
+        super().__init__(simulation_time, time_points)
+
+    def differential_eq(self, y, t, alpha, beta):
+        """In this function implements the differential equation of the SIR model will be implemented.
+
+        Args:
+            y (tuple): Vector that holds the current state of the three groups.
+            t (_): not used
+            alpha (_): not used
+            beta (_): not used
+
+        Returns:
+            tuple: Change amount for each group.
+        """
+        S, I, R = y
+        dSdt = -self.beta * ((S * I) / self.N) # -self.beta * S * I
+        dIdt = self.beta * ((S * I) / self.N) - self.alpha * I # self.beta * S * I - self.alpha * I
+        dRdt = self.alpha * I
+        return dSdt, dIdt, dRdt
+
+    def generate(self):
+        """This funtion generates the data for this configuration of the SIR model.
+        """
+        y_0 = self.S_0, self.I_0, self.R_0
+        self.data = odeint(self.differential_eq, y_0, self.t, args=(self.alpha, self.beta)).T
+        super().generate()
+
+    def plot(self):
+        """Plot the data which was generated.
+        """
+        super().plot(('Susceptible', 'Infectious', 'Removed'))
+
+    def save(self, name=''):
+        if self.generated:
+            COVID_Data = np.asarray([self.t, *self.data]) 
+
+            np.savetxt('datasets/SIR_data.csv', COVID_Data, delimiter=",")
+        else: 
+            print('Data has to be generated before plotting!')
+        
+
+class SIDR(SyntheticDeseaseData):
+    def __init__(self, N=59e6, I_0=1, D_0=0, R_0=0, simulation_time=500, time_points=100, alpha=0.191, beta=0.05, gamma=0.0294) -> None:
+        """This class is able to generate synthetic data for the SIDR model.
+
+        Args:
+            N (int, optional): Size of the population. Defaults to 59e6.
+            I_0 (int, optional): Initial size of the infectious group. Defaults to 1.
+            D_0 (int, optional): Initial size of the dead group. Defaults to 0.
+            R_0 (int, optional): Initial size of the recovered group. Defaults to 0.
+            simulation_time (int, optional): Real time for that the synthetic data is supposed to be generated in days. Defaults to 500.
+            time_points (int, optional): Number of time sample points. Defaults to 100.
+            alpha (float, optional): Factor dictating how many people per timestep go from 'Susceptible' to 'Infectious'. Defaults to 0.191.
+            beta (float, optional): Factor dictating how many people per timestep go from 'Infectious' to 'Dead'. Defaults to 0.05.
+            gamma (float, optional): Factor dictating how many people per timestep go from 'Infectious' to 'Recovered'. Defaults to 0.0294.
+        """
+        self.N = N
+        self.S_0 = N - I_0 - D_0 - R_0
+        self.I_0 = I_0
+        self.D_0 = D_0
+        self.R_0 = R_0
+
+        self.alpha = alpha
+        self.beta = beta
+        self.gamma = gamma
+
+        super().__init__(simulation_time, time_points)
+    
+    def differential_eq(self, y, t, alpha, beta, gamma):
+        """In this function implements the differential equation of the SIDR model will be implemented.
+
+        Args:
+            y (tuple): Vector that holds the current state of the three groups.
+            t (_): not used
+            alpha (_): not used
+            beta (_): not used
+            gamma (_): not used
+
+        Returns:
+            tuple: Change amount for each group.
+        """
+        S, I, D, R = y
+        dSdt = - (self.alpha / self.N) * S * I
+        dIdt = (self.alpha / self.N) * S * I - self.beta * I - self.gamma * I 
+        dDdt = self.gamma * I
+        dRdt = self.beta * I
+        return dSdt, dIdt, dDdt, dRdt
+
+    def generate(self):
+        """This funtion generates the data for this configuration of the SIR model.
+        """
+        y_0 = self.S_0, self.I_0, self.D_0, self.R_0
+        self.data = odeint(self.differential_eq, y_0, self.t, args=(self.alpha, self.beta, self.gamma)).T
+        super().generate()
+
+    def plot(self):
+        """Plot the data which was generated.
+        """
+        super().plot(('Susceptible', 'Infectious', 'Dead', 'Recovered'))
+
+    def save(self, name=''):
+        if self.generated:
+            COVID_Data = np.asarray([self.t, *self.data]) 
+
+            np.savetxt('datasets/SIDR_data.csv', COVID_Data, delimiter=",")
+        else: 
+            print('Data has to be generated before plotting!')