def build_transfers_timeline(df_opt_pre, df_opt_post, p1, p2, p3):
df_opt_post.loc[:, 'Date'] = pd.to_datetime(df_opt_post['Date'],
format='y%m%d').dt.date
timeline_cols = ["Date", "Shortage"]
df_opt_pre = df_opt_pre.loc[df_opt_pre.State == 'US']
df_opt_pre = df_opt_pre[timeline_cols]
no_model_visual = get_no_model_visual()
model_visual = get_model_visual()
df_opt_pre.columns = ["Date", no_model_visual["Shortage"]]
df_opt_post = df_opt_post.loc[
(df_opt_post.State == 'US') & \
(df_opt_post.Param1==float(p1)) & \
(df_opt_post.Param2==float(p2)) & \
(df_opt_post.Param3==float(p3))
]
df_opt_post = df_opt_post[timeline_cols]
df_opt_post.columns = ["Date", model_visual["Shortage"]]
df_opt_effect = pd.merge(df_opt_pre, df_opt_post, on='Date', how='inner')
return us_timeline(df_opt_effect, "Optimization Effect on Shortage", True)
def build_transfers_timeline(chosen_model, p1, p2, p3):
global df_mod1_shortages
global df_mod2_shortages
global df_mod1_projections
global df_mod2_projections
if chosen_model == "Washington IHME":
df_opt_pre = df_mod1_projections.copy()
df_opt_post = df_mod1_shortages.copy()
else:
df_opt_pre = df_mod2_projections.copy()
df_opt_post = df_mod2_shortages.copy()
timeline_cols = ["Date", "Shortage"]
df_opt_pre = df_opt_pre.loc[df_opt_pre.State == 'US']
df_opt_pre = df_opt_pre[timeline_cols]
df_opt_pre.columns = ["Date", no_model_visual["Shortage"]]
df_opt_post = df_opt_post.loc[
(df_opt_post.State == 'US') & \
(df_opt_post.Param1==float(p1)) & \
(df_opt_post.Param2==float(p2)) & \
(df_opt_post.Param3==float(p3))
]
df_opt_post = df_opt_post[timeline_cols]
df_opt_post.columns = ["Date", model_visual["Shortage"]]
df_opt_effect = pd.merge(df_opt_pre, df_opt_post, on='Date', how='inner')
return us_timeline(df_opt_effect, "Optimization Effect on Shortage", True)
def build_transfers_timeline(chosen_model,p1,p2,p3):
if chosen_model == "Washington IHME":
df_opt_pre, _ = get_df_mod1_projections()
df_opt_post = pd.read_csv('data/predicted_ventilator/state_supplies_table-ihme.csv', sep=",", parse_dates = ['Date'])
else:
df_opt_pre = get_df_mod2_projections()
df_opt_post = pd.read_csv('data/predicted_ventilator/state_supplies_table-ode.csv', sep=",", parse_dates = ['Date'])
df_opt_post.loc[:,'Date'] = pd.to_datetime(df_opt_post['Date'], format='y%m%d').dt.date
timeline_cols = ["Date","Shortage"]
df_opt_pre = df_opt_pre.loc[df_opt_pre.State == 'US']
df_opt_pre = df_opt_pre[timeline_cols]
no_model_visual = get_no_model_visual()
model_visual = get_model_visual()
df_opt_pre.columns = ["Date",no_model_visual["Shortage"]]
df_opt_post = df_opt_post.loc[
(df_opt_post.State == 'US') & \
(df_opt_post.Param1==float(p1)) & \
(df_opt_post.Param2==float(p2)) & \
(df_opt_post.Param3==float(p3))
]
df_opt_post = df_opt_post[timeline_cols]
df_opt_post.columns = ["Date",model_visual["Shortage"]]
df_opt_effect = pd.merge(df_opt_pre,df_opt_post,on='Date',how='inner')
return us_timeline(df_opt_effect,"Optimization Effect on Shortage",True)
def update_shortage_timeline(chosen_model):
if chosen_model == "Washington IHME":
df_projections_vent_us = df_mod1_projections
else:
df_projections_vent_us = df_mod2_projections
df_projections_vent_us = df_projections_vent_us.loc[df_projections_vent_us.State == 'US']
return us_timeline(df_projections_vent_us, "US Ventilator Supply, Demand, & Shortage", False)
def build_shortage_timeline(chosen_model):
global df_mod1_projections
global df_mod2_projections
if chosen_model == "Washington IHME":
df_projections_vent_us = df_mod1_projections.copy()
else:
df_projections_vent_us = df_mod2_projections.copy()
df_projections_vent_us = df_projections_vent_us.loc[df_projections_vent_us.State == 'US']
return us_timeline(df_projections_vent_us, "US Ventilator Supply, Demand, & Shortage", False)