# Parameters
country_ = "China"
end_date = "2020-04-01"
cad_start = "2020-01-22"
cad_end = "2020-03-16"
import pandas as pd
import numpy as np
import seaborn as sns; sns.set()
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
from sklearn.linear_model import LinearRegression
from scipy import stats
import statsmodels.api as sm
import pylab
# from google.colab import files
# from io import StringIO
# uploaded = files.upload()
url = 'https://raw.githubusercontent.com/assemzh/ProbProg-COVID-19/master/full_grouped.csv'
data = pd.read_csv(url)
data.Date = pd.to_datetime(data.Date)
# for fancy python printing
from IPython.display import Markdown, display
def printmd(string):
display(Markdown(string))
import warnings
warnings.filterwarnings('ignore')
import matplotlib as mpl
mpl.rcParams['figure.dpi'] = 250
# function to make the time series of confirmed and daily confirmed cases for a specific country
def create_country (country, end_date, state = False) :
if state :
df = data.loc[data["Province/State"] == country, ["Province/State", "Date", "Confirmed", "Deaths", "Recovered"]]
else :
df = data.loc[data["Country/Region"] == country, ["Country/Region", "Date", "Confirmed", "Deaths", "Recovered"]]
df.columns = ["country", "date", "confirmed", "deaths", "recovered"]
# group by country and date, sum(confirmed, deaths, recovered). do this because countries have multiple cities
df = df.groupby(['country','date'])['confirmed', 'deaths', 'recovered'].sum().reset_index()
# convert date string to datetime
std_dateparser = lambda x: str(x)[5:10]
df.date = pd.to_datetime(df.date)
df['date_only'] = df.date.apply(std_dateparser)
df = df.sort_values(by = "date")
df = df[df.date <= end_date]
# make new confirmed cases every day:
cases_shifted = np.array([0] + list(df.confirmed[:-1]))
daily_confirmed = np.array(df.confirmed) - cases_shifted
df["daily_confirmed"] = daily_confirmed
fig, ax = plt.subplots(nrows=1, ncols=1, figsize=(8, 6))
ax = [ax]
sns.lineplot(x = df.date,
y = df.daily_confirmed,
ax = ax[0])
ax[0].set(ylabel='Daily Confirmed Cases')
ax[0].axvline(pd.to_datetime('2020-01-22'),
linestyle = '--', linewidth = 1.5,
label = "Policy start: Jan 22, 2020" ,
color = "red")
ax[0].xaxis.get_label().set_fontsize(22)
ax[0].yaxis.get_label().set_fontsize(22)
x = df.date
# ax[0].set_xticks(x[::30])
ax[0].xaxis.set_major_locator(mdates.MonthLocator(interval=1)) #to get a tick every month
ax[0].title.set_fontsize(20)
ax[0].tick_params(labelsize=22)
myFmt = mdates.DateFormatter('%b %-d, %Y')
ax[0].xaxis.set_major_formatter(myFmt)
ax[0].set(ylabel='', xlabel='');
ax[0].legend(loc = "bottom right", fontsize=22)
sns.set_style("ticks")
plt.tight_layout()
sns.despine()
plt.savefig('/content/sample_data/country_daily.pdf')
print(df.tail())
return df
def summary(samples):
site_stats = {}
for k, v in samples.items():
site_stats[k] = {
"mean": torch.mean(v, 0),
"std": torch.std(v, 0),
"5%": v.kthvalue(int(len(v) * 0.05), dim=0)[0],
"95%": v.kthvalue(int(len(v) * 0.95), dim=0)[0],
}
return site_stats
cad = create_country(country_, end_date)
cad = cad[cad.date >= cad_start].reset_index(drop = True)
cad["days_since_start"] = np.arange(cad.shape[0]) + 1
cad.shape
cad_tmp = cad[cad.date < cad_end]
cad_tmp.shape
reg_data = cad_tmp.copy()
reg_data.head()
!pip install pyro-ppl
!pip install numpyro
import torch
import pyro
import pyro.distributions as dist
from torch import nn
from pyro.nn import PyroModule, PyroSample
from pyro.infer import MCMC, NUTS, HMC
from pyro.infer.autoguide import AutoGuide, AutoDiagonalNormal
from pyro.infer import SVI, Trace_ELBO
from pyro.infer import Predictive
# we should be able to have an empirical estimate for the mean of the prior for the 2nd regression bias term
# this will be something like b = log(max(daily_confirmed))
# might be able to have 1 regression model but change the data so that we have new terms for (tau < t)
# like an interaction term
class COVID_change(PyroModule):
def __init__(self, in_features, out_features, b1_mu, b2_mu):
super().__init__()
self.linear1 = PyroModule[nn.Linear](in_features, out_features, bias = False)
self.linear1.weight = PyroSample(dist.Normal(0.5, 0.25).expand([1, 1]).to_event(1))
self.linear1.bias = PyroSample(dist.Normal(b1_mu, 1.))
# could possibly have stronger priors for the 2nd regression line, because we wont have as much data
self.linear2 = PyroModule[nn.Linear](in_features, out_features, bias = False)
self.linear2.weight = PyroSample(dist.Normal(0., 0.25).expand([1, 1])) #.to_event(1))
self.linear2.bias = PyroSample(dist.Normal(b2_mu, b2_mu/4))
def forward(self, x, y=None):
tau = pyro.sample("tau", dist.Beta(4, 3))
sigma = pyro.sample("sigma", dist.Uniform(0., 3.))
# fit lm's to data based on tau
sep = int(np.ceil(tau.detach().numpy() * len(x)))
mean1 = self.linear1(x[:sep]).squeeze(-1)
mean2 = self.linear2(x[sep:]).squeeze(-1)
mean = torch.cat((mean1, mean2))
obs = pyro.sample("obs", dist.StudentT(2, mean, sigma), obs=y)
return mean
tensor_data = torch.tensor(reg_data[["confirmed", "days_since_start", "daily_confirmed"]].values, dtype=torch.float)
x_data = tensor_data[:, 1].unsqueeze_(1)
y_data = np.log(tensor_data[:, 0])
y_data_daily = np.log(tensor_data[:, 2])
# prior hyper params
# take log of the average of the 1st quartile to get the prior mean for the bias of the 2nd regression line
q1 = np.quantile(y_data, q = 0.25)
bias_1_mean = np.mean(y_data.numpy()[y_data <= q1])
print("Prior mean for Bias 1: ", bias_1_mean)
# take log of the average of the 4th quartile to get the prior mean for the bias of the 2nd regression line
q4 = np.quantile(y_data, q = 0.75)
bias_2_mean = np.mean(y_data.numpy()[y_data >= q4])
print("Prior mean for Bias 2: ", bias_2_mean)
model = COVID_change(1, 1,
b1_mu = bias_1_mean,
b2_mu = bias_2_mean)
# need more than 400 samples/chain if we want to use a flat prior on b_2 and w_2
num_samples = 400
# mcmc
nuts_kernel = NUTS(model)
mcmc = MCMC(nuts_kernel,
num_samples=num_samples,
warmup_steps = 200,
num_chains = 1)
mcmc.run(x_data, y_data)
samples = mcmc.get_samples()
# extract individual posteriors
weight_1_post = samples["linear1.weight"].detach().numpy()
weight_2_post = samples["linear2.weight"].detach().numpy()
bias_1_post = samples["linear1.bias"].detach().numpy()
bias_2_post = samples["linear2.bias"].detach().numpy()
tau_post = samples["tau"].detach().numpy()
sigma_post = samples["sigma"].detach().numpy()
# build likelihood distribution:
tau_days = list(map(int, np.ceil(tau_post * len(x_data))))
mean_ = torch.zeros(len(tau_days), len(x_data))
obs_ = torch.zeros(len(tau_days), len(x_data))
for i in range(len(tau_days)) :
mean_[i, :] = torch.cat((x_data[:tau_days[i]] * weight_1_post[i] + bias_1_post[i],
x_data[tau_days[i]:] * weight_2_post[i] + bias_2_post[i])).reshape(len(x_data))
obs_[i, :] = dist.Normal(mean_[i, :], sigma_post[i]).sample()
samples["_RETURN"] = mean_
samples["obs"] = obs_
pred_summary = summary(samples)
mu = pred_summary["_RETURN"] # mean
y = pred_summary["obs"] # samples from likelihood: mu + sigma
y_shift = np.exp(y["mean"]) - np.exp(torch.cat((y["mean"][0:1], y["mean"][:-1])))
print(y_shift)
predictions = pd.DataFrame({
"days_since_start": x_data[:, 0],
"mu_mean": mu["mean"], # mean of likelihood
"mu_perc_5": mu["5%"],
"mu_perc_95": mu["95%"],
"y_mean": y["mean"], # mean of likelihood + noise
"y_perc_5": y["5%"],
"y_perc_95": y["95%"],
"true_confirmed": y_data,
"true_daily_confirmed": y_data_daily,
"y_daily_mean": y_shift
})
w1_ = pred_summary["linear1.weight"]
w2_ = pred_summary["linear2.weight"]
b1_ = pred_summary["linear1.bias"]
b2_ = pred_summary["linear2.bias"]
tau_ = pred_summary["tau"]
sigma_ = pred_summary["sigma"]
ind = int(np.ceil(tau_["mean"] * len(x_data)))
mcmc.summary()
diag = mcmc.diagnostics()
print(ind)
print(reg_data.date[ind])
sns.distplot(weight_1_post,
kde_kws = {"label": "Weight posterior before CP"},
# color = "red",
norm_hist = True,
kde = True)
plt.axvline(x = w1_["mean"], linestyle = '--',label = "Mean weight before CP" ,
# color = "red"
)
sns.distplot(weight_2_post,
kde_kws = {"label": "Weight posterior after CP"},
color = "red",
norm_hist = True,
kde = True)
plt.axvline(x = w2_["mean"], linestyle = '--',label = "Mean weight after CP" ,
color = "red")
legend = plt.legend(loc='upper right')
legend.get_frame().set_alpha(1)
sns.set_style("ticks")
plt.tight_layout()
sns.despine()
plt.savefig('/content/sample_data/country_weights.pdf')
print(w1_["mean"])
print(w2_["mean"])
predictions['date'] = pd.to_datetime(reg_data.date)
start_date_ = str(reg_data.date[0]).split(' ')[0]
change_date_ = str(reg_data.date[ind]).split(' ')[0]
print("Date of change for {}: {}".format(country_, change_date_))
import seaborn as sns
# plot data:
fig, ax = plt.subplots(nrows=1, ncols=1, figsize=(7, 5))
ax = [ax]
# log regression model
ax[0].scatter(y = np.exp(y_data[:ind]), x = x_data[:ind], s = 15);
ax[0].scatter(y = np.exp(y_data[ind:]), x = x_data[ind:], s = 15, color = "red");
ax[0].plot(predictions["days_since_start"],
np.exp(predictions["y_mean"]),
color = "green",
label = "Fitted line by MCMC-NUTS model")
# ax[0].axvline(2,
# linestyle = '--', linewidth = 1.5,
# label = "Date of Lockdown: Jan 23, 2020" ,
# color = "red")
# ax[0].axvline(ind,
# linestyle = '--', linewidth = 1.5,
# label = "Date of Change: Feb 8, 2020",
# color = "black")
ax[0].fill_between(predictions["days_since_start"],
np.exp(predictions["y_perc_5"]),
np.exp(predictions["y_perc_95"]),
alpha = 0.25,
label = "90% CI of predictions",
color = "teal");
ax[0].fill_betweenx([0, 1],
tau_["5%"] * len(x_data),
tau_["95%"] * len(x_data),
alpha = 0.25,
label = "90% CI of changing point",
color = "lightcoral",
transform=ax[0].get_xaxis_transform());
ax[0].set(ylabel = "Total Cases",)
ax[0].legend(loc = "lower right", fontsize=12.8)
ax[0].set_ylim([100,150000])
ax[0].xaxis.get_label().set_fontsize(16)
ax[0].yaxis.get_label().set_fontsize(16)
ax[0].title.set_fontsize(20)
ax[0].tick_params(labelsize=16)
# plt.xticks(ticks=[1,17,34,51], labels=["Jan 22",
# "Feb 8",
# "Feb 25",
# "Mar 13"], fontsize=15)
ax[0].set_yscale('log')
plt.setp(ax[0].get_xticklabels(), rotation=0, horizontalalignment='center')
print(reg_data.columns)
myFmt = mdates.DateFormatter('%m-%d')
sns.set_style("ticks")
sns.despine()
ax[0].figure.savefig('/content/sample_data/country_cp.pdf')