def update_id_output(n_clicks, value):
if n_clicks > 0:
if value:
samples = sf.master_df()
samples.drop_duplicates(inplace=True)
parsed_samples = [x.strip() for x in value.split(",")]
status_df = samples[samples['run_sample_id'].isin(parsed_samples)].iloc[:, [0,1,3,11,10,15,19,17,18]]
status_df['accession_date'] = status_df['accession_date'] + ', ' + \
pd.to_datetime(
status_df['accession_date']).dt.strftime('%a')
sample_status = dash_table.DataTable(columns=[{"name": i, "id": i} for i in status_df],
data=status_df.to_dict(
'records'),
style_data_conditional=[
{
'if': {'row_index': 'odd'},
'backgroundColor': 'rgb(248,248,248)'
}
],
style_table={'height': min(40 + 35 * (len(status_df)), 500),
'overflowY': 'auto'}
)
return sample_status
def per_detect(fc_click, start_date, end_date, value):
fig = go.Figure()
if fc_click and value:
x_lower = (dt.strptime(start_date, '%Y-%m-%d') -
timedelta(days=1)).strftime('%m-%d-%Y')
x_upper = (dt.strptime(end_date, '%Y-%m-%d') +
timedelta(days=1)).strftime('%m-%d-%Y')
samples = sf.master_df()
g_samples = samples[(samples['date'] >= pd.to_datetime(start_date))
& (samples['date'] <= pd.to_datetime(end_date))]
g_samples.loc[:, 'call_map'] = g_samples.call.map({
'not_detected': 0,
'detected': 1
}) * 100
percent_detect = g_samples[g_samples['Site Name'].isin(value)].groupby(
['Site Name', 'date']).mean().reset_index().rename(
columns={'call_map': 'Percent Detected'})
fig = {}
try:
title = "Percent Detected By Site From " + start_date[:
10] + " To " + end_date[:
10]
plot_title = {'text': [title], 'fontSize': 16}
chart = alt.Chart(percent_detect).mark_line().encode(
x=alt.X('date',
title='Date',
axis=alt.Axis(titleFontSize=14,
labelFontSize=12,
formatType='time',
format='%a, %b-%d',
labelAngle=45,
grid=False),
scale=alt.Scale(domain=[x_lower, x_upper])),
y=alt.Y(
'Percent Detected',
title='Percent Detected',
scale=alt.Scale(domain=[
0, max(percent_detect['Percent Detected']) * 1.05
])),
tooltip=['date', 'Percent Detected', 'Site Name'],
color='Site Name')
fig = chart.properties(
title=plot_title, width=1200,
height=250).configure_title(anchor='middle').to_dict()
except Exception:
pass
return fig
def pie_chart(n_clicks):
samples = sf.master_df()
percent = samples[samples['call'] == 'detected'].dropna(
subset=['Site Name']).replace(
{'Healing Grove Health Center': 'Healing Grove'})
low_sites = percent.groupby('Site Name').count()[['runid']].reset_index()
low_sites['runid'] = (low_sites['runid'] /
low_sites.runid.sum()).mul(100).round(2)
low_sites = low_sites.query("runid < 1")['Site Name'].values.tolist()
percent.loc[percent['Site Name'].isin(low_sites), 'Site Name'] = "Other"
percent = percent['Site Name'].value_counts(
normalize=True).mul(100).round(2)
data = go.Pie(labels=percent.index.tolist(),
values=percent.values.tolist())
layout = go.Layout(annotations=[
dict(xref='paper',
yref='paper',
x=0.5,
y=-0.22,
showarrow=False,
text=
f'Other Testing Sites: {*low_sites[0:3],} <br> {*low_sites[3:],}')
],
title={
'text': '% Positive Detection By Site',
'y': 0.9,
'x': 0.5,
'xanchor': 'center',
'yanchor': 'top'
})
fig = go.Figure(data=data, layout=layout)
fig.update_traces(textinfo='percent+label',
marker=dict(line=dict(color='#000000', width=2)),
textposition='outside')
return fig
def update_query_output(n_clicks, start_date, end_date, value):
if n_clicks > 0:
samples = sf.master_df()
samples.drop_duplicates(inplace=True)
samples = samples[samples['Specimen receipt date'] != 0]
if value and not start_date and not end_date:
status_df = samples[samples['Site Name'].isin(value)].sort_values(by='date', ascending=False) \
.iloc[:, [0,1,3,11,10,15,19,17,18]]
elif start_date and end_date and not value:
status_df = samples[(samples['date'] >= pd.to_datetime(start_date)) & (
samples['date'] <= pd.to_datetime(end_date))] \
.iloc[:, [0,1,3,11,10,15,19,17,18]]
elif start_date and end_date and value:
status_df = samples[(samples['date'] >= pd.to_datetime(start_date)) & (
samples['date'] <= pd.to_datetime(end_date))]
status_df = status_df[status_df['Site Name'].isin(value)].sort_values(by='date', ascending=False) \
.iloc[:, [0,1,3,11,10,15,19,17,18]]
else:
return
status_df['accession_date'] = status_df['accession_date'] + ', ' + \
pd.to_datetime(
status_df['accession_date']).dt.strftime('%a')
sample_status = dash_table.DataTable(columns=[{"name": i, "id": i} for i in status_df],
data=status_df.to_dict('records'),
style_data_conditional=[
{
'if': {'row_index': 'odd'},
'backgroundColor': 'rgb(248,248,248)'
}
],
style_table={'height': 500,
'overflowY': 'auto'},
sort_action='native'
)
return sample_status
def sample_status(fc_click, start_date, end_date):
fig = {}
sample_status = []
if fc_click:
x_lower = (dt.strptime(start_date, '%Y-%m-%d') -
timedelta(days=1)).strftime('%m-%d-%Y')
x_upper = (dt.strptime(end_date, '%Y-%m-%d') +
timedelta(days=1)).strftime('%m-%d-%Y')
samples = sf.master_df()
g_samples = samples[(samples['date'] >= pd.to_datetime(start_date))
& (samples['date'] <= pd.to_datetime(end_date))]
dummies = pd.get_dummies(g_samples.flags)
result = pd.concat([g_samples, dummies], axis=1)
result = result.groupby('date').mean().reset_index()
wout_pass = np.delete(g_samples.flags.unique(),
np.where(g_samples.flags.unique() == 'pass'))
for i in wout_pass:
result[i] = (result[i] * 100).round(2)
result.rename(columns={i: 'percent-' + i}, inplace=True)
status = pd.wide_to_long(result,
'percent',
i="date",
j="condition",
sep='-',
suffix=r'.*').reset_index()
title = "Percent Flag Occurrence From " + start_date[:
10] + " To " + end_date[:
10]
plot_title = {'text': [title], 'fontSize': 16}
chart = alt.Chart(status).mark_line().encode(
x=alt.X('date',
title='Date',
axis=alt.Axis(titleFontSize=14,
labelFontSize=12,
formatType='time',
format='%a, %b-%d',
labelAngle=45,
grid=False),
scale=alt.Scale(domain=[x_lower, x_upper])),
y=alt.Y('percent',
title='Percent Occurrence',
scale=alt.Scale(domain=[0, max(status.percent) * 1.05])),
tooltip=['date', 'percent', 'condition'],
color='condition')
annotation = alt.Chart(status).mark_text(
align='center', fontSize=15,
dy=-10).encode(x='date', y='percent',
text='percent').transform_filter(
(datum.percent > 0))
fig = (chart + annotation).properties(
title=plot_title, width=1200,
height=250).configure_title(anchor='middle').to_dict()
status_df = g_samples[g_samples['flags'] != 'pass'] \
.sort_values(by='date', ascending=False).iloc[:, [0, 1, 3, 11, 15, 19, 17, 18]]
status_df['accession_date'] = status_df['accession_date'] + ', ' + \
pd.to_datetime(
status_df['accession_date']).dt.strftime('%a')
sample_status = dash_table.DataTable(columns=[{
"name": i,
"id": i
} for i in status_df],
data=status_df.to_dict('records'),
style_data_conditional=[{
'if': {
'row_index': 'odd'
},
'backgroundColor':
'rgb(248,248,248)'
}],
style_table={
'height': '450px',
'overflowY': 'auto'
},
sort_action='native')
return fig, sample_status
def samples_plot(n_clicks, start_date, end_date):
fig_total = {}
fig_tat = {}
if n_clicks:
x_lower = (dt.strptime(start_date, '%Y-%m-%d') -
timedelta(days=1)).strftime('%m-%d-%Y')
x_upper = (dt.strptime(end_date, '%Y-%m-%d') +
timedelta(days=1)).strftime('%m-%d-%Y')
samples = sf.master_df()
samples = samples[(samples['date'] >= pd.to_datetime(start_date))
& (samples['date'] <= pd.to_datetime(end_date))]
#### Total Sample Plot ####
total_samples = samples.copy()
total_samples.loc[:, 'sample_type'].replace({'Sample': 1},
inplace=True)
total = total_samples.groupby(
'date').sum().loc[:, ['sample_type']].reset_index()
# make fig
title = "Total Clinical Samples per Day, " + start_date[:10] + " to " + end_date[:10] + ": " + str(
total['sample_type'].sum()) + " Samples"
plot_title = {'text': [title], 'fontSize': 16}
bars = alt.Chart(total).mark_bar(size=15).encode(
x=alt.X('date:T',
title='Date',
axis=alt.Axis(titleFontSize=14,
labelFontSize=12,
formatType='time',
format='%a, %b-%d',
labelAngle=45,
grid=False),
scale=alt.Scale(domain=[x_lower, x_upper])),
y=alt.Y(
'sample_type:Q',
axis=alt.Axis(labelFontSize=12,
titleFontSize=14,
title='Number of Samples',
domain=False),
scale=alt.Scale(domain=[0, max(total['sample_type']) * 1.1])),
tooltip=['date', 'sample_type'])
bar_text = bars.mark_text(
baseline='top',
align='center',
dy=-15,
color='black',
fontSize=13,
).encode(text=alt.Text('sample_type:Q'))
fig_total = (bars + bar_text).properties(
title=plot_title, width=1200, height=250).configure_title(
anchor='middle').interactive().to_dict()
#### Turnaround Time Plot ####
tat_samples = samples.copy()
tat_samples.dropna(subset=['AIMS Reporting Notified Date'],
inplace=True)
time_diff = pd.to_datetime(tat_samples['AIMS Reporting Notified Date']) \
- pd.to_datetime(tat_samples['accession_date'])
tat_samples['Turnaround Time in Hours'] = (
time_diff.dt.total_seconds() / 3600)
tat_samples = tat_samples.groupby('date').agg({
'Turnaround Time in Hours':
'mean'
}).astype('int').reset_index()
# make fig
title = "Turnaround Time of Samples per Day, " + start_date[:10] + " to " + end_date[
:10] + " Average Turnaround Time: " + \
str(tat_samples['Turnaround Time in Hours'].mean().astype('int')) + " Hours"
plot_title = {'text': [title], 'fontSize': 16}
bars = alt.Chart(tat_samples).mark_bar(size=15).encode(
x=alt.X('date:T',
title='Date',
axis=alt.Axis(titleFontSize=14,
labelFontSize=12,
formatType='time',
format='%a, %b-%d',
labelAngle=45,
grid=False),
scale=alt.Scale(domain=[x_lower, x_upper])),
y=alt.Y(
'Turnaround Time in Hours:Q',
axis=alt.Axis(labelFontSize=12,
titleFontSize=14,
title='Number of Samples',
domain=False),
scale=alt.Scale(domain=[
0, max(tat_samples['Turnaround Time in Hours']) * 1.1
])),
tooltip=['date', 'Turnaround Time in Hours'])
bar_text = bars.mark_text(
baseline='top',
align='center',
dy=-15,
color='black',
fontSize=13,
).encode(text=alt.Text('Turnaround Time in Hours:Q'))
rule = alt.Chart(tat_samples).mark_rule(
color='red', opacity=0.5,
strokeDash=[3, 3]).encode(y='mean(Turnaround Time in Hours):Q')
fig_tat = (bars + bar_text + rule).properties(
title=plot_title, width=1200, height=250).configure_title(
anchor='middle').interactive().to_dict()
return fig_total, fig_tat
def get_layout():
master = sf.master_df()
positive = master[master['call'] == 'detected']
time_diff = pd.to_datetime(master['AIMS Reporting Notified Date']) \
- pd.to_datetime(master['accession_date'])
tat = (time_diff.dropna().dt.total_seconds() / 3600).mean().astype(int)
heading = sf.make_row(children=[
sf.make_col(children=[
html.H2(f"To this day, we have processed {len(master)} samples."),
html.Div(
html.H2(f"Of those samples, {len(positive)} are positive.")),
html.Div(
html.
H2(f"Average Turnaround Time: {tat} Hours from Accessioning to Reporting."
))
]),
sf.make_col(
dcc.Graph(id='location-pie',
style={'height': 500},
config={'displayModeBar': False}))
])
date_range = html.Div([
dcc.DatePickerRange(id='date-picker-range'),
html.Div(id='output-container-date-picker-range')
])
fc_submit = sf.make_row(children=[
sf.make_col(html.Button("Submit",
id="fc-button-clinical",
n_clicks=0,
n_clicks_timestamp=0),
col_num=1)
])
submit_res = html.Div(id='button-res-clinical')
site = pd.read_csv(
'/ghds/groups/labdesk/bshih/c19dash/c19_dashboard/c19_tat.csv').fillna(
0)
site = site[site['Site Name'].notnull()]['Site Name'].unique()
site.sort()
site_dropdown = html.Div(
dcc.Dropdown(id='percent-detect-site-dropdown',
options=[{
"label": i,
"value": i
} for i in site],
placeholder='Collection Site',
multi=True))
tot_smp_plot = html.Div([dav.VegaLite(id='tot-smp-plot')],
style={
'width': '100%',
'height': '400px'
})
tat_plot = html.Div([dav.VegaLite(id='tat-plot')],
style={
'width': '100%',
'height': '400px'
})
percent_detect = html.Div([dav.VegaLite(id='percent-detect-plot')],
style={
'width': '100%',
'height': '400px'
})
total_sample = html.Div([dav.VegaLite(id='total-sample-plot')],
style={
'width': '100%',
'height': '400px'
})
smp_status_plot = html.Div([dav.VegaLite(id='smp-status-plot')],
style={
'width': '100%',
'height': '400px'
})
smp_status_table = html.Div(id='smp-status-table',
style={
'width': '100%',
'overflowX': 'auto'
})
layout = html.Div([
heading,
html.H3('Pick a date range '), date_range, fc_submit, submit_res,
html.H3('Sample Statistics'), tot_smp_plot, tat_plot,
html.H3('Site Specific Metrics'), site_dropdown, percent_detect,
total_sample,
html.H3('Sample Status'), smp_status_plot, smp_status_table
],
style={'width': '100%'})
return layout