/
aux_files_data_exploration.py
199 lines (180 loc) · 8.34 KB
/
aux_files_data_exploration.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
import pandas as pd
import numpy as np
import seaborn as sns
sns.set(font_scale=2)
import matplotlib.pyplot as plt
pd.plotting.register_matplotlib_converters()
from ipywidgets import interact,Layout,interactive_output,SelectMultiple,HBox,Label
from scipy.optimize import curve_fit
from sklearn.metrics import r2_score
import warnings
warnings.filterwarnings("ignore")
df=pd.read_csv('merged_us_data_by_state_with_mobility_and_policy.csv',index_col=0,parse_dates=[1]).reset_index(drop=True)
df=df.rename(columns={'retail_and_recreation_percent_change_from_baseline': 'retail and recreation',
'grocery_and_pharmacy_percent_change_from_baseline': 'grocery and pharmacy',
'parks_percent_change_from_baseline': 'parks',
'transit_stations_percent_change_from_baseline': 'transit',
'workplaces_percent_change_from_baseline': 'workplaces',
'residential_percent_change_from_baseline': 'residential'})
df=df[~df.state.isin(['PR','AS','MP','VI','GU'])]
states=df.state.unique()
state_tuples_by_infected=[]
state_tuples_by_deaths=[]
for state in states:
total_infected=df[df.state==state]['positive'].max()
total_deaths=df[df.state==state]['death'].max()
state_tuples_by_infected.append((total_infected,state))
state_tuples_by_deaths.append((total_deaths,state))
state_tuples_by_infected.sort(reverse=True)
state_tuples_by_deaths.sort(reverse=True)
states_by_infected=[c[1] for c in state_tuples_by_infected]
states_by_deaths=[c[1] for c in state_tuples_by_deaths]
national_df=df.groupby('date').sum()
def align_dates(df,thr=100):
df=df.sort_values(by='date').reset_index()
ind_thr=(df['positive']>=thr).idxmax()
ref_date=df.loc[ind_thr,'date']
df['date']=(df['date']-ref_date)
df['date']=df['date'].apply(lambda x: x.days)
return df
first=True
for state in states:
if first:
df_aligned=align_dates(df[df.state==state])
first=False
else:
df_aligned=pd.concat([df_aligned,align_dates(df[df.state==state])],axis=0)
df_aligned.reset_index(inplace=True)
def plot_vs_events(state,col,aligned=False,logy=False,logx=False,events=False,ylabel='',mvavg=1):
if aligned:
mydf=df_aligned[df.state==state]
else:
mydf=df[df.state==state]
mydf=mydf.reset_index(drop=True)
plt.plot(mydf.date[mvavg-1:],moving_average(mydf[col],n=mvavg),label=col)
if logx:
plt.xscale('log')
if logy:
plt.yscale('log')
if events:
if mydf['pandemic declared'].idxmax()>0:
plt.axvline(mydf.loc[mydf['pandemic declared'].idxmax(),'date'],
label='pandemic declared',linestyle='--',color='g')
if mydf['school canceled'].idxmax()>0:
plt.axvline(mydf.loc[mydf['school canceled'].idxmax(),'date'],
label='school canceled',linestyle='-.',color='y')
if mydf['stay at home'].idxmax()>0:
plt.axvline(mydf.loc[mydf['stay at home'].idxmax(),'date'],
label='stay at home',linestyle=':',color='r')
#'stay at home', 'pandemic declared','school canceled'
if aligned:
plt.xlabel('days since 100th case')
else:
plt.ylabel('date')
plt.xticks(rotation=90)
if len(ylabel)>0:
plt.ylabel(ylabel)
def plot_mobility(state,aligned,mvavg=1):
plt.figure(figsize=(12,6))
plt.title('Change in mobility for {}'.format(state))
plot_vs_events(state,col='retail and recreation',aligned=aligned,mvavg=mvavg)
plot_vs_events(state,col='grocery and pharmacy',aligned=aligned,mvavg=mvavg)
#plot_vs_events(state,col='parks',aligned=True,mvavg=mvavg)
plot_vs_events(state,col='transit',aligned=aligned,mvavg=mvavg)
plot_vs_events(state,col='workplaces',aligned=aligned,mvavg=mvavg)
plot_vs_events(state,col='residential',aligned=aligned,mvavg=mvavg,logy=False,
events=True,ylabel='percent change\n relative to baseline')
plt.legend(bbox_to_anchor=(1.04,1), loc='upper left', ncol=1)
plt.show()
def make_widgets_mobility():
style = {'description_width': 'initial','width': 100}
select=SelectMultiple(
options=states_by_infected,
value=[states_by_infected[0]],
style=style,
disabled=False
)
select=HBox([Label('Select a state (or multiple states):'), select])
display(select)
return select
def fit_and_plot_test_results():
tvec=np.array((national_df.index-national_df.index[0]).days)
logistic = lambda x,a,b,c: c/(1+np.exp(-a*(x-b)))
plt.figure(figsize=(12,6))
popt, pcov = curve_fit(logistic, tvec,national_df['totalTestResultsIncrease'].values)
logistic_parameters=popt
plt.title('Nationwide test totals')
plt.plot(tvec,national_df['totalTestResultsIncrease'],label='tests administered')
R2=r2_score(y_true=national_df['totalTestResultsIncrease'].values,
y_pred=logistic(tvec, *popt))
plt.plot(tvec, logistic(tvec, *popt), 'r-', label='logistic fit ($R^2$={:.3f})'.format(R2))
plt.plot(tvec,popt[2]+0*tvec,label='daily testing capacity \n estimate: {}'.format(int(np.round(popt[2]))),color='k',linestyle='--')
plt.xlabel('days since {}/{}'.format(national_df.index[0].month,national_df.index[0].day))
plt.legend(bbox_to_anchor=(1.04,1), loc='upper left', ncol=1)
plt.show()
def make_widgets_testing():
from ipywidgets import interact,Layout,interactive_output,SelectMultiple,HBox,Label,IntSlider
style = {'description_width': 'initial','width': 100}
select=SelectMultiple(
options=states_by_infected,
value=[states_by_infected[0]],
style=style,
disabled=False
)
select=HBox([Label('Select a state (or multiple states):'), select])
mov_avg=IntSlider(
value=1,
min=1,
max=14,
step=1,
orientation='horizontal',
readout=True,
readout_format='d'
)
mov_avg=HBox([Label('Select a window size. Values larger than one use a moving average to smooth the curve:'), mov_avg])
display(select)
display(mov_avg)
return select,mov_avg
def moving_average(a, n=3) :
a=pd.Series(a).rolling(window=n).mean().iloc[n-1:].values
return a
def plot_test_results(state,n):
my_df=df[df.state==state]
my_df=my_df.set_index('date')
tvec=np.array((my_df.index-my_df.index[0]).days)
yvec=my_df['totalTestResultsIncrease'].values
plt.figure(figsize=(12,6))
plt.title('Tests for {}'.format(state))
plt.plot(moving_average(tvec,n=n),moving_average(yvec,n=n),label='tests administered')
plt.xlabel('days since {}/{}'.format(my_df.index[0].month,national_df.index[0].day))
#plt.legend(bbox_to_anchor=(1.04,1), loc='upper left', ncol=1)
plt.show()
def plot_positive_rate(window=1):
fig,ax=plt.subplots(2,1,figsize=(12,8),sharex=True)
ax[0].plot(national_df.index[window-1:],moving_average(national_df['positiveIncrease'],n=window)/moving_average(national_df['totalTestResultsIncrease'],n=window),label='% positive')
ax[1].plot(national_df.index[window-1:],moving_average(national_df['totalTestResultsIncrease'],n=window),label='tests administered')
ax[1].plot(national_df.index[window-1:],moving_average(national_df['positiveIncrease'],n=window),label='positive tests')
ax[1].set_xlabel('date')
ax[0].set_ylabel('% positive')
#ax[1].set_ylabel('tests administered')
ax[0].set_ylim(0,1.1)
ax[1].legend()
plt.xticks(rotation=90)
plt.show()
def plot_positive_rate_by_state(state,window):
my_df=df[df.state==state]
my_df=my_df.set_index('date')
tvec=np.array((my_df.index-my_df.index[0]).days)
yvec=my_df['totalTestResultsIncrease'].values
fig,ax=plt.subplots(2,1,figsize=(12,8),sharex=True)
ax[0].set_title('Tests for {}'.format(state))
ax[0].plot(my_df.index[window-1:],moving_average(my_df['positiveIncrease'],n=window)/moving_average(my_df['totalTestResultsIncrease'],n=window),label='% positive')
ax[1].plot(my_df.index[window-1:],moving_average(my_df['totalTestResultsIncrease'],n=window),label='tests administered')
ax[1].plot(my_df.index[window-1:],moving_average(my_df['positiveIncrease'],n=window),label='positive tests')
ax[1].set_xlabel('date')
ax[0].set_ylabel('% positive')
#ax[1].set_ylabel('tests administered')
ax[0].set_ylim(0,1.1)
ax[1].legend()
plt.xticks(rotation=90)
plt.show()