g = sns.FacetGrid(df, col='winby')
g.map(plt.hist, 'R_Age', bins=20)
sns.lmplot(x="B__Round1_Strikes_Body Significant Strikes_Attempts",
y="B__Round1_Strikes_Body Significant Strikes_Landed",
col="winner",
hue="winner",
data=df,
col_wrap=2,
size=6)
cnt_srs = df['R_Location'].value_counts().head(15)
trace = go.Bar(
x=cnt_srs.index,
y=cnt_srs.values,
marker=dict(color=cnt_srs.values, ),
)
layout = go.Layout(title='Most Popular cities for Red fighters')
data = [trace]
fig = go.Figure(data=data, layout=layout)
offline.iplot(fig, filename="Ratio")
cnt_srs = df['B_Location'].value_counts().head(15)
trace = go.Bar(
x=cnt_srs.index,
y=cnt_srs.values,
marker=dict(color=cnt_srs.values, ),
def plot_evaluation_metrics(self,
metric_dict: Dict,
config_names: List = None,
xlabel: str = None,
ylabel: str = "Metric value",
title: str = None,
showlegend: bool = True):
"""Returns a barplot of the train and test values of ``metric_dict`` of ``config_names``.
Value of a metric for all ``config_names`` are plotted as a grouped bar.
Train and test values of a metric are plot side-by-side for easy comparison.
Parameters
----------
metric_dict : `dict` [`str`, `callable`]
Evaluation metrics to compute. Same as
`~greykite.framework.framework.benchmark.benchmark_class.BenchmarkForecastConfig.get_evaluation_metrics`.
To get the best visualization, keep number of metrics <= 2.
config_names : `list` [`str`], default None
Which config results to plot. A list of config names.
If None, uses all the available config keys.
xlabel : `str` or None, default None
x-axis label.
ylabel : `str` or None, default "Metric value"
y-axis label.
title : `str` or None, default None
Plot title.
showlegend : `bool`, default True
Whether to show the legend.
Returns
-------
fig : `plotly.graph_objs.Figure`
Interactive plotly bar plot.
"""
evaluation_metrics_df = self.get_evaluation_metrics(
metric_dict=metric_dict, config_names=config_names)
# This function groups by config name
evaluation_metrics_df = (evaluation_metrics_df.drop(
columns=["split_num"]).groupby("config_name").mean().dropna(
how="all"))
# Rearranges columns so that train and test error of a config are side by side
plot_df = pd.DataFrame()
for metric_name in metric_dict.keys():
plot_df[f"train_{metric_name}"] = evaluation_metrics_df[
f"train_{metric_name}"]
plot_df[f"test_{metric_name}"] = evaluation_metrics_df[
f"test_{metric_name}"]
if title is None:
title = "Average evaluation metric across rolling windows"
data = []
# Each row (index) is a config. Adds each row to the bar plot.
for index in plot_df.index:
data.append(
go.Bar(name=index,
x=plot_df.columns,
y=plot_df.loc[index].values))
layout = go.Layout(
xaxis=dict(title=xlabel),
yaxis=dict(title=ylabel),
title=title,
showlegend=showlegend,
barmode="group",
)
fig = go.Figure(data=data, layout=layout)
return fig
with open('trendtweets.csv', 'w', newline='') as trendfile: #puts trend list items in CSV
writeto = csv.writer(trendfile, delimiter=',')
writeto.writerows(trendlistexport)
xtrends = []
trendvolumes = []
for item in trendlistexport:
xvalue = item[0]
xtrends.append(xvalue)
for item in trendlistexport:
yvalue = item[1]
trendvolumes.append(yvalue)
del xtrends[0]
del trendvolumes[0]
data = [go.Bar(
x=xtrends,
y=trendvolumes)]
data = data
layout = go.Layout(
title='Trending Topic vs. Volume of Tweets',
xaxis=dict(
title='Trending Topic',
titlefont=dict(
family='Courier New, monospace',
size=18,
color='#7f7f7f'
)
),
yaxis=dict(
title='Volume of Tweets',
titlefont=dict(
from plotly import graph_objs
from plotly.offline import download_plotlyjs, init_notebook_mode, plot, iplot
init_notebook_mode(connected=True)
# In[31]:
get_ipython().magic('matplotlib inline')
df = raw_data.airline_sentiment
neg = len(df[raw_data["airline_sentiment"] == "negative"])
pos = len(df[raw_data["airline_sentiment"] == "positive"])
neu = len(df[raw_data["airline_sentiment"] == "neutral"])
dist = [
graph_objs.Bar(
x=["negative","neutral","positive"],
y=[neg, neu, pos],
)]
plotly.offline.iplot({"data":dist, "layout":graph_objs.Layout(title="Sentiment type distribution in Airline dataset")})
# In[33]:
get_ipython().magic('matplotlib inline')
# Plot histogram of tweets per airline
airline_count = Counter(raw_data['airline2'])
ac_df = pd.DataFrame.from_dict(airline_count, orient = 'index')
ac_df.plot(kind='bar')
u_name = pd.DataFrame(fake__users.name.str.split(' ', 1).tolist(),
columns=['first', 'last'])
user_name = u_name.groupby('first',
as_index=False).size().reset_index(name='counts')
user_name = user_name.sort_values('counts', ascending=False).head(20)
# bar plot
#first names
first_name = u_name.groupby('first',
as_index=False).size().reset_index(name='counts')
first_name = first_name.sort_values('counts', ascending=False).head(20)
df = go.Bar(x=first_name['counts'],
y=first_name['first'],
orientation='h',
name='First Name')
#last names
first_name = u_name.groupby('last',
as_index=False).size().reset_index(name='counts')
first_name = first_name.sort_values('counts', ascending=False).head(20)
df1 = go.Bar(x=first_name['counts'],
y=first_name['last'],
orientation='h',
name='Last Name')
fig = tools.make_subplots(rows=1,
cols=2,
subplot_titles=('first Name', 'Last Name'))
fig.append_trace(df, 1, 1)
fig.append_trace(df1, 1, 2)
html.Div([
dcc.Markdown('''Shopping'''),
sliders[5],
], style={"width": "15em", "padding": "1em", "display": "inline-block"}),
html.Div([
dcc.Markdown('''Geld'''),
sliders[6],
], style={"width": "15em", "padding": "1em", "display": "inline-block"}),
html.Div([
html.Div([
dcc.Graph(
id = "detail_view_of_pois",
figure = {
"data": [
go.Bar(
x = ["Bildung","Gesundheit", "Freizeit", "Shopping", "Geld", "Öffentliche Gebäude", "Gastwirtschaft"],
y = [0,0,0,0,0,0,0])
],
"layout" : go.Layout(
title = "POIs",
yaxis = dict(
tick0 = 0,
dtick = 1)
)
}
)
], style={"display": "inline-block"}),
html.Div([
dcc.Graph(
id = "housing_distribution",
figure = {
html.Div([
html.Div([
html.H5("Feature Comparisons with Gender"),
html.P("These graphs represent a comparison between the Gender and several other features presented in the data set. ", style={'color': 'gray', 'fontSize': 14})
], className="twelve columns"),
], className="row"),
html.Div([
html.Div([
dcc.Graph(
figure=go.Figure(
data=[
go.Bar(
x=gender_disabled_df['Gender'],
y=gender_disabled_df['Disabled'],
name='Disabled',
marker=go.bar.Marker(
color='rgb(225, 128, 0)'
)
),
go.Bar(
x=gender_not_disabled_df['Gender'],
y=gender_not_disabled_df['Disabled'],
name='Not Disabled',
marker=go.bar.Marker(
color='rgb(55, 83, 109)'
)
)
],
layout=go.Layout(
title='Gender vs Disability',
showlegend=False,
yaxis=dict(title='NIV'))
fig = go.Figure(data, layout)
py.iplot(fig)
# -
# Average age at onset when NIV is used:
ALS_proc_df[ALS_proc_df.NIV == 1]['Age at onset'].mean()
# +
configure_plotly_browser_state()
ALS_proc_NIV_3R = ALS_proc_df.groupby(['3R', 'NIV'
]).REF.count().to_frame().reset_index()
data = [
go.Bar(x=ALS_proc_NIV_3R[ALS_proc_NIV_3R.NIV == 0]['3R'],
y=ALS_proc_NIV_3R[ALS_proc_NIV_3R.NIV == 0]['REF'],
name='Not used'),
go.Bar(x=ALS_proc_NIV_3R[ALS_proc_NIV_3R.NIV == 1]['3R'],
y=ALS_proc_NIV_3R[ALS_proc_NIV_3R.NIV == 1]['REF'],
name='Using NIV')
]
layout = go.Layout(barmode='group')
fig = go.Figure(data=data, layout=layout)
py.iplot(fig, filename='grouped-bar')
# -
# Average 3R value when NIV is used:
ALS_proc_df[ALS_proc_df.NIV == 1]['3R'].mean()
# **Comments:** Clearly, there's a big dependence of the use of NIV with the respiratory symptoms indicated by 3R, as expected.
def layout():
indicadores = json.load(open('indicadores.json'))
datasets = indicadores['datasets']
# ==============================
# Barras por publicador
# ==============================
publicadores = indicadores['datasets_por_publicador']
publicadores_yaxis = []
for x in publicadores:
publicadores_yaxis.append('<br>'.join(wrap(x['_id']['publicador'])))
barras_publicador = dcc.Graph(figure=go.Figure(
data=[
go.Bar(
y=publicadores_yaxis,
x=[x['datasets'] for x in publicadores],
orientation='h',
)
],
layout=go.Layout(height=6000,
margin=go.layout.Margin(l=600),
yaxis={
'categoryorder': 'array',
'categoryarray': publicadores_yaxis
})), )
# ==============================
# Barras por fuente
# ==============================
fuentes = indicadores['datasets_por_fuente']
fuentes_yaxis = []
for x in fuentes:
fuentes_yaxis.append('<br>'.join(wrap(x['_id']['fuente'])))
barras_fuentes = dcc.Graph(figure=go.Figure(
data=[
go.Bar(
y=fuentes_yaxis,
x=[x['datasets'] for x in fuentes],
orientation='h',
)
],
layout=go.Layout(height=6000,
margin=go.layout.Margin(l=600),
yaxis={
'categoryorder': 'array',
'categoryarray': fuentes_yaxis
})), )
# ==============================
# Barras por frecuencia
# ==============================
frecuencias = indicadores['datasets_por_frecuencia']
barras_frecuencias = dcc.Graph(figure=go.Figure(data=[
go.Bar(y=[x['_id']['frecuencia'] for x in frecuencias],
x=[x['datasets'] for x in frecuencias],
orientation='h')
]))
# ==============================
# Barras por formato
# ==============================
formatos = indicadores['recursos_por_formato']
barras_formatos = dcc.Graph(figure=go.Figure(data=[
go.Bar(y=[x['_id']['formato'] for x in formatos],
x=[x['recursos'] for x in formatos],
orientation='h')
]))
# ==============================
# Torta por actualización
# ==============================
actualizados = len([
x for x in datasets if x['dias'] <= dias_frecuencias[x['frecuencia']]
])
desactualizados = len(
[x for x in datasets if x['dias'] > dias_frecuencias[x['frecuencia']]])
torta_actualizacion = dcc.Graph(
id='graph',
figure=go.Figure(data=[
go.Pie(
values=[actualizados, desactualizados],
labels=['Actualizados', 'Desactualizados'],
)
]))
# ==============================
# Tabla desactualizados
# ==============================
datasets = [
x for x in datasets if x['dias'] > dias_frecuencias[x['frecuencia']]
]
datasets = sorted(datasets, key=lambda x: x['dias'])[::-1][:10]
datasets = [{
'Titulo': d['titulo'],
'Frecuencia de actualización': d['frecuencia'],
'Días desactualizado': int(d['dias']),
'Fuente': d['fuente'],
'Publicador': d['publicador'],
'URL de descarga': d['url']
} for d in datasets]
tabla_datasets = dash_table.DataTable(id='table',
columns=[{
"name": i,
"id": i
} for i in datasets[0].keys()],
data=datasets,
style_table={'overflowX': 'scroll'})
# ==============================
# Nube de keywords
# ==============================
keywords = {
x['_id']['keyword']: x['datasets']
for x in indicadores['datasets_por_keyword']
}
buffered_keywords = BytesIO()
WordCloud(width=500, height=700,
background_color='white').generate_from_frequencies(
keywords).to_image().save(buffered_keywords, format="JPEG")
img_str_keywords = base64.b64encode(buffered_keywords.getvalue())
nube_keywords = html.Img(
src="data:image/jpeg;base64,{}".format(img_str_keywords.decode()))
# ==============================
# Nube de búsquedas
# ==============================
busquedas = {
x['_id']['keyword']: x['datasets']
for x in indicadores['datasets_por_keyword']
}
buffered_busquedas = BytesIO()
WordCloud(width=500, height=700,
background_color='white').generate_from_frequencies(
busquedas).to_image().save(buffered_busquedas, format="JPEG")
img_str_busquedas = base64.b64encode(buffered_busquedas.getvalue())
nube_busquedas = html.Img(
src="data:image/jpeg;base64,{}".format(img_str_busquedas.decode()))
return html.Div([
html.Div([
html.Div([
html.H3('Actualización', className='text-center'),
torta_actualizacion,
],
className='col-xs-4 text-center'),
html.Div([
html.H3('Frecuencias', className='text-center'),
barras_frecuencias,
],
className='col-xs-4 text-center'),
html.Div([
html.H3('Formatos', className='text-center'),
barras_formatos,
],
className='col-xs-4 text-center'),
],
className='row'),
html.H3('Top 10 datasets desactualizados', className='text-center'),
tabla_datasets,
html.Div([
html.Div(
[html.H3('Keywords', className='text-center'), nube_keywords],
className='col-xs-6 text-center'),
html.Div([
html.H3('Búsquedas', className='text-center'), nube_busquedas
],
className='col-xs-6 text-center'),
],
className='row margin-bottom-xl'),
html.Div([
html.Div([
html.H3('Publicadores', className='text-center'),
barras_publicador,
],
className='col-xs-6 text-center'),
html.Div([
html.H3('Fuentes', className='text-center'),
barras_fuentes,
],
className='col-xs-6 text-center'),
],
className='row')
],
className='container')
def update_agg_mode_bar_chart(selected_ucr, selected_neighbourhood,
selected_agg_mode, selected_n, start_date,
end_date):
df = crime_main_geo[(crime_main_geo['UCR'].isin(selected_ucr)) &
(crime_main_geo['Name'].isin(selected_neighbourhood))]
df = df.rename(columns={'d1': 'DATE'})
df = df[(df['DATE'] >= start_date) & (df['DATE'] <= end_date)]
# AGGREGATION OF CRIME REPORTS
df_grouped = df.groupby(by=selected_agg_mode).count()[["CODE"
]].reset_index()
df_grouped.columns = [selected_agg_mode, 'COUNT']
fig_agg_mode = go.Figure(
go.Bar(x=df_grouped[selected_agg_mode],
y=df_grouped['COUNT'],
marker_color='lightsalmon'))
fig_agg_mode.update_xaxes(type='category')
fig_agg_mode.update_layout(height=390,
yaxis_title="Number of Crime Reports")
# TOP CRIME CATEGORIES
df_top = df.value_counts('OFFENSE_CODE_GROUP').to_frame().reset_index()
df_top.columns = ['OFFENSE_CODE_GROUP', 'COUNT']
df_top = df_top.head(selected_n)
fig_top_crimes = go.Figure(
go.Bar(x=df_top['COUNT'],
y=df_top['OFFENSE_CODE_GROUP'],
orientation='h',
marker_color='indianred'))
fig_top_crimes.update_layout(height=400, yaxis=dict(autorange="reversed"))
# PROPORTION OF CRIME CLASSES
colors_1 = ['#E0BBE4', '#957DAD', '#D291BC', '#FEC8D8']
df_donut_crime_classes = df["CRIME_CLASS"].value_counts().to_frame(
).reset_index()
df_donut_crime_classes.columns = ["CRIME_CLASS", 'COUNT']
fig_donut_crime_class = go.Figure(data=[
go.Pie(labels=df_donut_crime_classes["CRIME_CLASS"],
values=df_donut_crime_classes['COUNT'],
hole=0.4)
])
fig_donut_crime_class.update_traces(marker=dict(colors=colors_1))
fig_donut_crime_class.update_layout(legend=dict(
orientation="v", yanchor="bottom", y=-0.3, xanchor="right", x=1))
fig_donut_crime_class.update_layout(margin=dict(t=20, b=20, l=20, r=20))
# PROPORTION OF SHOOTING INCIDENTS
colors_2 = ["#85DE77", "#FF756D"]
df_donut_shooting = df["SHOOTING"].value_counts().to_frame().reset_index()
df_donut_shooting.columns = ["SHOOTING", 'COUNT']
fig_donut_shooting = go.Figure(data=[
go.Pie(labels=df_donut_shooting["SHOOTING"],
values=df_donut_shooting['COUNT'],
hole=0.4)
])
fig_donut_shooting.update_traces(marker=dict(colors=colors_2))
fig_donut_shooting.update_layout(legend=dict(
orientation="v", yanchor="bottom", y=-0.07, xanchor="right", x=1))
fig_donut_shooting.update_layout(margin=dict(t=0, b=0, l=0, r=0))
# CHOROPLETH MAP OF CRIME COUNTS BY GEOGRAPHICAL REGIONS
df_choro = df.groupby(by="GEOID10").count()[['OFFENSE_CODE_GROUP'
]].reset_index()
df_choro['GEOID10'] = df_choro['GEOID10'].astype('str')
df_choro = df_choro[df_choro['GEOID10'] != '0']
df_choro_merged = pd.merge(boston_polygon,
df_choro,
how='left',
on='GEOID10')
fig_choro = go.Figure(
go.Choroplethmapbox(geojson=boston_geo_ok,
locations=df_choro_merged['GEOID10'],
z=df_choro_merged['OFFENSE_CODE_GROUP'],
colorscale="Blues",
featureidkey="GEOID10",
marker_opacity=0.5,
marker_line_width=0.7))
fig_choro.update_layout(mapbox_style="carto-positron",
mapbox_zoom=10,
mapbox_center={
"lat": 42.33,
"lon": -71.09
})
fig_choro.update_layout(margin={
"r": 0,
"t": 0,
"l": 0,
"b": 0
},
height=930)
fig_choro.update_geos(fitbounds="locations", visible=True)
return fig_agg_mode, fig_top_crimes, fig_donut_crime_class, fig_donut_shooting, fig_choro
def updateGraph(dataFields:list, filterIndex:int, graphType:int, binSize:int): """ updates the graph based on the chosen data fields, data filters, graph type, and bin size (the latter if histogram is selected) """ # title of the graph, set to the filename for now title = 'Without filter' if len(dataFields) == 0: return go.Figure(layout=dict(title=title)) # empty graph if filterIndex is 0: fList = ['isMale'] elif filterIndex is 1: fList = ['analyticMajor'] elif filterIndex is 2: fList = ['nativeEnglish'] # function which filters a piece of data # depending on the filters the user selected def dataFilter(data:dict) -> bool: for name in fList: return True filteredDataSet = tuple(filter(dataFilter, dataSet)) if len(filteredDataSet) == 0: return go.Figure(layout=dict(title=title)) # empty graph # convert the data being plotted into numbers try: traceValues = [ [ float(d[field]) for d in filteredDataSet ] for field in dataFields ] except ValueError: return go.Figure(layout=dict(title="Error: Can't plot non-numeric data on a numeric axis.")) # turn the position on the graph type slider into a graph type name graphType = GRAPHTYPE_CHOICES[graphType] if graphType == 'Histogram': out = ff.create_distplot( traceValues, dataFields, show_curve=False, show_rug=False, bin_size=binSize, ) out.layout['title'] = title return out if graphType == 'Density Plot': out = ff.create_distplot( traceValues, dataFields, show_hist=False, show_rug=False, ) out.layout['title'] = title return out layout = dict(title=title) # layout used by all of the graph types below if graphType == 'Violin Plot': traces = [ dict( type='violin', name=field, y=values, ) for field,values in zip(dataFields,traceValues) ] elif graphType == 'Box Plot': traces = [ go.Box( name=field, y=values, ) for field,values in zip(dataFields,traceValues) ] elif graphType == 'Dot Plot': traces = [ dict( type='scatter', name=field, y=[d[field] for d in filteredDataSet], x=[d[DATA_IDFIELD] for d in filteredDataSet], mode='markers', ) for field in dataFields ] layout['xaxis'] = dict( title=DATA_IDFIELD, type='category', titlefont=dict( size=12, ), ) elif graphType == 'Bar Plot': traces = [ go.Bar( name=field, y=[d[field] for d in filteredDataSet], x=[d[DATA_IDFIELD] for d in filteredDataSet], ) for field in dataFields ] layout['xaxis'] = dict( title=DATA_IDFIELD, type='category', titlefont=dict( size=12, ), ) return go.Figure(data=traces, layout=layout)
# color_continuous_scale=["#FFF1A8", "#FFD608"],
range_color=(min(df_proov_cum['Pct acumulado proveedores']),
max(df_proov_cum['Pct acumulado proveedores'])),
mapbox_style="carto-positron",
zoom=4,
center={
"lat": 4.570868,
"lon": -74.2973328
},
opacity=0.5)
fig_map_2.update_layout(margin={"r": 0, "t": 0, "l": 0, "b": 0})
# 2. Bar Plot
trace_1 = go.Bar(x=df_proov_cum['Pct acumulado proveedores'],
y=df_proov_cum['Departamento'],
orientation='h')
layout = go.Layout(hovermode='closest')
fig = go.Figure(data=[trace_1], layout=layout) #
fig.update_layout(font=dict(color="#252525", family="Roboto"),
paper_bgcolor='rgba(0,0,0,0)',
plot_bgcolor='rgba(0,0,0,0)',
xaxis_title="% concentracion 10 primeros contratistas",
yaxis=dict(dtick=1),
margin={
"r": 0,
"t": 0,
"l": 0,
"b": 0
})
def update_plots(n_clicks, start_date, end_date, show_exponential,
normalise_by_pop, align_cases_check, align_cases_input,
align_deaths_check, align_deaths_input,
align_active_cases_check, align_active_cases_input,
align_daily_cases_check, align_daily_cases_input,
align_daily_deaths_check, align_daily_deaths_input,
saved_json_data, *args):
print(n_clicks, start_date, end_date, args)
start_date = datetime.datetime.strptime(start_date, '%Y-%m-%d').date()
end_date = datetime.datetime.strptime(end_date, '%Y-%m-%d').date()
country_names = []
for country in args:
country_names.extend(country)
if saved_json_data is None:
country_data = {}
else:
country_data = json.loads(saved_json_data)
for i, country in enumerate(country_names):
if country not in country_data.keys():
try:
data = get_data(country)
country_data[country] = data
except Exception as e:
print(e)
country_names.remove(country)
continue
out = []
for title in [
'Cases', 'Deaths', 'Currently Infected', 'Daily New Cases',
'Daily New Deaths'
]:
if normalise_by_pop:
axis_title = f"{title} (% of population)"
else:
axis_title = title
if title == 'Cases':
align_countries = align_cases_check
align_input = align_cases_input
elif title == 'Deaths':
align_countries = align_deaths_check
align_input = align_deaths_input
elif title == 'Currently Infected':
align_countries = align_active_cases_check
align_input = align_active_cases_input
elif title == 'Daily New Cases':
align_countries = align_daily_cases_check
align_input = align_daily_cases_input
elif title == 'Daily New Deaths':
align_countries = align_daily_deaths_check
align_input = align_daily_deaths_input
figs = []
if align_countries:
xaxis_title = f'Days since the total confirmed cases reached {align_input}'
if normalise_by_pop:
xaxis_title += '% of the population'
else:
xaxis_title = ''
layout_normal = {
'yaxis': {
'title': axis_title,
'type': 'linear',
'showgrid': True
},
'xaxis': {
'title': xaxis_title,
'showgrid': True
},
'showlegend':
True,
'margin': {
'l': 70,
'b': 100,
't': 0,
'r': 0
},
'updatemenus': [
dict(buttons=list([
dict(args=[
"yaxis", {
'title': axis_title,
'type': 'linear',
'showgrid': True
}
],
label="Linear",
method="relayout"),
dict(args=[
"yaxis", {
'title': axis_title,
'type': 'log',
'showgrid': True
}
],
label="Logarithmic",
method="relayout")
]),
direction="down",
pad={
"r": 10,
"t": 10,
"b": 10
},
showactive=True,
x=0.,
xanchor="left",
y=1.2,
yanchor="top"),
]
}
# if normalise_by_pop:
# layout_normal['yaxis']['tickformat'] = '%.2f'
layout_daily_plot = copy.deepcopy(layout_normal)
layout_daily_plot['updatemenus'].append(
dict(buttons=list([
dict(args=[{
"visible":
[False, False] + [False, False, True] * len(country_names)
if show_exponential else [False] +
[False, True] * len(country_names)
}],
label="Bar",
method="update"),
dict(args=[{
"visible":
[True, True] + [True, True, False] * len(country_names)
if show_exponential else [True] +
[True, False] * len(country_names)
}],
label="Scatter",
method="update")
]),
direction="down",
pad={
"r": 10,
"t": 10,
"b": 10
},
showactive=True,
x=0.2,
xanchor="left",
y=1.2,
yanchor="top"), )
if show_exponential:
figs.append(
go.Scatter(
x=[datetime.date(2020, 2, 20)]
if not align_countries else [0],
y=[0],
mode='lines',
line={
'color': 'black',
'dash': 'dash'
},
showlegend=True,
visible=False if title in [
'Daily New Cases', 'Daily New Deaths'
] else 'legendonly',
name=fr'Best exponential fits',
yaxis='y1',
legendgroup='group2',
))
label = fr'COUNTRY : best fit (doubling time)'
else:
label = fr'COUNTRY'
figs.append(
go.Scatter(
x=[datetime.date(2020, 2, 20)] if not align_countries else [0],
y=[0],
mode='lines+markers',
line={'color': 'black'},
showlegend=True,
visible=False if title
in ['Daily New Cases', 'Daily New Deaths'] else 'legendonly',
name=label,
yaxis='y1',
legendgroup='group2',
))
for i, c in enumerate(country_names):
print(c)
if country_data[c] is None:
print("Cannot retrieve data from country:", c)
continue
if title == 'Daily New Cases':
dates = country_data[c]['Cases']['dates'][1:]
xdata = np.arange(len(dates))
ydata = np.diff(
np.array(country_data[c]['Cases']['data']).astype('float'))
elif title == 'Daily New Deaths':
dates = country_data[c]['Deaths']['dates'][1:]
xdata = np.arange(len(dates))
ydata = np.diff(
np.array(
country_data[c]['Deaths']['data']).astype('float'))
elif title not in country_data[c]:
continue
else:
dates = country_data[c][title]['dates']
xdata = np.arange(len(dates))
ydata = country_data[c][title]['data']
ydata = np.array(ydata).astype('float')
date_objects = []
for date in dates:
date_objects.append(
datetime.datetime.strptime(date, '%Y-%m-%d').date())
date_objects = np.asarray(date_objects)
if normalise_by_pop:
ydata = ydata / POPULATIONS[c] * 100
if align_countries:
if title in ['Cases', 'Deaths']:
idx_when_n_cases = np.abs(ydata - align_input).argmin()
elif title in ['Currently Infected', 'Daily New Cases']:
ydata_cases = np.array(
country_data[c]['Cases']['data']).astype('float')
ydata_cases = ydata_cases / POPULATIONS[
c] * 100 if normalise_by_pop else ydata_cases
idx_when_n_cases = np.abs(ydata_cases -
align_input).argmin()
elif title in ['Daily New Deaths']:
ydata_cases = np.array(
country_data[c]['Deaths']['data']).astype('float')
ydata_cases = ydata_cases / POPULATIONS[
c] * 100 if normalise_by_pop else ydata_cases
idx_when_n_cases = np.abs(ydata_cases -
align_input).argmin()
if title in ['Daily New Cases', 'Daily New Deaths']:
idx_when_n_cases -= 1
xdata = xdata - idx_when_n_cases
model_date_mask = (date_objects <= end_date) & (date_objects >=
start_date)
model_dates = []
model_xdata = []
date = start_date
d_idx = min(xdata[model_date_mask])
while date <= end_date:
model_dates.append(date)
model_xdata.append(d_idx)
date += datetime.timedelta(days=1)
d_idx += 1
model_xdata = np.array(model_xdata)
b, logA = np.polyfit(xdata[model_date_mask],
np.log(ydata[model_date_mask]), 1)
lin_yfit = np.exp(logA) * np.exp(b * model_xdata)
if show_exponential:
if np.log(2) / b > 1000 or np.log(2) / b < 0:
double_time = 'no growth'
else:
double_time = fr'{np.log(2) / b:.1f} days to double'
label = fr'{c.upper():<10s}: {np.exp(b):.2f}^t ({double_time})'
else:
label = fr'{c.upper():<10s}'
figs.append(
go.Scatter(
x=date_objects if not align_countries else xdata,
y=ydata,
hovertext=[
f"Date: {d.strftime('%d-%b-%Y')}" for d in date_objects
] if align_countries else '',
mode='lines+markers',
marker={'color': colours[i]},
line={'color': colours[i]},
showlegend=True,
visible=False if title
in ['Daily New Cases', 'Daily New Deaths'] else True,
name=label,
yaxis='y1',
legendgroup='group1',
))
if show_exponential:
if np.log(2) / b < 0:
show_plot = False
else:
show_plot = True
figs.append(
go.Scatter(
x=model_dates if not align_countries else model_xdata,
y=lin_yfit,
hovertext=[
f"Date: {d.strftime('%d-%b-%Y')}"
for d in model_dates
] if align_countries else '',
mode='lines',
line={
'color': colours[i],
'dash': 'dash'
},
showlegend=False,
visible=False if title in [
'Daily New Cases', 'Daily New Deaths'
] else show_plot,
name=fr'Model {c.upper():<10s}',
yaxis='y1',
legendgroup='group1',
))
if title in ['Daily New Cases', 'Daily New Deaths']:
figs.append(
go.Bar(x=date_objects if not align_countries else xdata,
y=ydata,
hovertext=[
f"Date: {d.strftime('%d-%b-%Y')}"
for d in date_objects
] if align_countries else '',
showlegend=True,
visible=True,
name=label,
marker={'color': colours[i]},
yaxis='y1',
legendgroup='group1'))
layout_out = copy.deepcopy(layout_daily_plot)
else:
layout_out = copy.deepcopy(layout_normal)
out.append({'data': figs, 'layout': layout_out})
# Plot 'New Cases vs Total Cases' and 'New Deaths vs Total Deaths'
for title in ['Cases', 'Deaths']:
fig_new_vs_total = []
for i, c in enumerate(country_names):
l = 7 # Number of days to look back
cases = np.array(country_data[c][title]['data']).astype('float')
xdata = np.copy(cases[l:])
ydata = np.diff(cases)
len_ydata = len(ydata)
# Compute new cases over the past l days
ydata = np.sum([
np.array(ydata[i:i + l])
for i in range(len_ydata) if i <= (len_ydata - l)
],
axis=1)
dates = country_data[c][title]['dates'][l:]
date_objects = []
for date in dates:
date_objects.append(
datetime.datetime.strptime(date, '%Y-%m-%d').date())
date_objects = np.asarray(date_objects)
mask = xdata > 100 if title == 'Cases' else xdata > 10
xdata = xdata[mask]
ydata = ydata[mask]
date_objects = date_objects[mask]
if normalise_by_pop:
xdata = xdata / POPULATIONS[c] * 100
ydata = ydata / POPULATIONS[c] * 100
fig_new_vs_total.append(
go.Scatter(
x=xdata,
y=ydata,
hovertext=[
f"Date: {d.strftime('%d-%b-%Y')}" for d in date_objects
],
mode='lines+markers',
marker={'color': colours[i]},
line={'color': colours[i]},
showlegend=True,
name=fr'{c.upper():<10s}',
yaxis='y1',
legendgroup='group1',
))
if normalise_by_pop:
yaxis_title = f'New {title} (% of population) per week (log scale)' # {l} days'
xaxis_title = f'Total {title} (% of population) (log scale)'
else:
yaxis_title = f'New {title} per week' # {l} days)'
xaxis_title = f'Total {title}'
layout_new_vs_total = {
'yaxis': {
'title': yaxis_title,
'type': 'log',
'showgrid': True
},
'xaxis': {
'title': xaxis_title,
'type': 'log',
'showgrid': True
},
'showlegend': True,
'margin': {
'l': 70,
'b': 100,
't': 50,
'r': 0
},
}
out.append({'data': fig_new_vs_total, 'layout': layout_new_vs_total})
out.append(json.dumps(country_data))
out.append(None)
return out
WH_S_2012=WH_S[WH_S['Year']==2012]
WH_S_2012=pd.DataFrame(WH_S.groupby('Product_Category', as_index=False)['Order_Demand'].mean())
WH_S_2012= WH_S_2012.sort_values('Order_Demand', ascending=False)
WH_S_2013=WH_S[WH_S['Year']==2013]
WH_S_2013=pd.DataFrame(WH_S_2013.groupby('Product_Category', as_index=False)['Order_Demand'].mean())
WH_S_2013=WH_S_2013.sort_values('Order_Demand', ascending=False)
WH_S_2014=WH_S[WH_S['Year']==2014]
WH_S_2014=pd.DataFrame(WH_S_2014.groupby('Product_Category', as_index=False)['Order_Demand'].mean())
WH_S_2014=WH_S_2014.sort_values('Order_Demand', ascending=False)
WH_S_2015=WH_S[WH_S['Year']==2015]
WH_S_2015=pd.DataFrame(WH_S_2015.groupby('Product_Category', as_index=False)['Order_Demand'].mean())
WH_S_2015=WH_S_2015.sort_values('Order_Demand', ascending=False)
WH_S_2016=WH_S[WH_S['Year']==2016]
WH_S_2016=pd.DataFrame(WH_S_2016.groupby('Product_Category', as_index=False)['Order_Demand'].mean())
WH_S_2016=WH_S_2016.sort_values('Order_Demand', ascending=False)
trace1 = go.Bar(x=WH_S_2012['Product_Category'], y=WH_S_2012['Order_Demand'], name='Year_2012')
trace2 = go.Bar(x=WH_S_2013['Product_Category'], y=WH_S_2013['Order_Demand'], name='Year_2013')
trace3 = go.Bar(x=WH_S_2014['Product_Category'], y=WH_S_2014['Order_Demand'], name='Year_2014')
trace4 = go.Bar(x=WH_S_2015['Product_Category'], y=WH_S_2015['Order_Demand'], name='Year_2015')
trace5 = go.Bar(x=WH_S_2016['Product_Category'], y=WH_S_2016['Order_Demand'], name='Year_2016')
fig = tools.make_subplots(rows=2, cols=5)
fig.append_trace(trace5, 1, 1)
fig.append_trace(trace4, 1, 2)
fig.append_trace(trace3, 1, 3)
fig.append_trace(trace2, 1, 4)
fig.append_trace(trace1, 1, 5)
layout=fig['layout'].update(height=500, width=1200, title='Order demand vs product category with respect to all years for '+ str (Warehouse[i]),xaxis=dict(
title='Product Category',
titlefont=dict(
family='Courier New, monospace',
size=18,
def PCA_sumplot(Z,
Z_chose,
Theta_record,
pca_obj,
fig_dens_I=[],
new_data=[],
Ncols=2,
PC_select=2,
height=600,
width=1000):
titles = [
'probab', 'Ave. PC coordinates among kde sampled theta vectors',
'loadings of PC {}'.format(PC_select + 1)
]
fig_subplots = tools.make_subplots(rows=int(len(titles) / float(Ncols)) +
(len(titles) % Ncols > 0),
cols=Ncols,
subplot_titles=tuple(titles))
for gp in range(len(titles)):
pos1 = int(float(gp) / Ncols) + 1
pos2 = gp - (pos1 - 1) * Ncols + 1
title = titles[gp]
if gp == 0:
zprime = Z[Z_chose]
bandwidth = estimate_bandwidth(zprime, quantile=0.2, n_samples=500)
X_plot = np.linspace(-2, 8, 100)[:, np.newaxis]
kde_plot = KernelDensity(kernel='gaussian',
bandwidth=bandwidth).fit(zprime)
log_dens = kde_plot.score_samples(X_plot)
trace = go.Scatter(x=X_plot[:, 0],
y=np.exp(log_dens),
mode='lines',
fill='tozeroy',
line=dict(color='red', width=2))
fig_subplots.append_trace(trace, pos1, pos2)
if len(fig_dens_I):
fig_subplots.append_trace(fig_dens_I[0], pos1, pos2)
if gp == 1:
feat_sum = np.sum(new_data, axis=0)
trace = go.Bar(
x=['PC {}'.format(x + 1) for x in range(new_data.shape[1])],
y=feat_sum,
marker=dict(color='rgb(0,0,205)'))
fig_subplots.append_trace(trace, pos1, pos2)
fig_subplots['layout']['yaxis' + str(gp + 1)].update(title='mean')
if gp == 2:
times_data = [
list(Theta_record[x]['comb'][:, 0])
for x in Theta_record.keys()
]
times_data = np.array(times_data)
times_av = np.mean(times_data, axis=0)
times_av = [int(x) for x in times_av / 1000]
times_av = ['{}k y'.format(x) for x in times_av]
Xcomps = pca_obj.components_
trace = go.Bar(x=times_av,
y=Xcomps[PC_select, :],
marker=dict(color='rgb(0,0,205)'))
fig_subplots.append_trace(trace, pos1, pos2)
fig_subplots['layout']['yaxis' +
str(gp + 1)].update(title='eigen value')
fig_subplots['layout'].update(height=height, width=width)
layout = go.Layout(title=title)
fig = go.Figure(data=fig_subplots, layout=layout)
iplot(fig_subplots)
temp=app_train['NAME_CONTRACT_TYPE'].value_counts()
x=temp.index
y=temp.values
plt.pie(x=temp.values,explode=(0.1,0),labels=temp.index,startangle=80,autopct='%1.1f%%',
colors=['#F1BF1B','#B1F11B'],frame=False,radius=1.5)
temp = app_train['NAME_INCOME_TYPE'].value_counts()
temp_y0 = []
temp_y1=[]
for val in temp.index:
temp_y1.append(np.sum(app_train["TARGET"][app_train["NAME_INCOME_TYPE"]==val] == 1))
temp_y0.append(np.sum(app_train["TARGET"][app_train["NAME_INCOME_TYPE"]==val] == 1))
trace1 = go.Bar(x= temp.index, y = (temp_y1 / temp.sum()) * 100,name='YES') #go is plotly.graph_objs
trace2 = go.Bar(x= temp.index, y = (temp_y0 / temp.sum()) * 100,name='NO') #go is plotly.graph_objs
data = [trace1, trace2]
layout = go.Layout(
title = "Income sources of Applicant's in terms of loan is repayed or not in %",
#barmode='stack',
width = 1000,
xaxis=dict(
title='Income source',
tickfont=dict(
size=14,
color='rgb(107, 107, 107)'
)
),
yaxis=dict(
def impact_on_phenotype(mo_uncorrected_file, mo_corrected_file, fdrth=0.05, saveto='./impact_on_phenotype'):
"""
:param fdrth: fdr cutoff
:param saveto: path to save file
:param mo_uncorrected_file: mageck uncorrected output
:param mo_corrected_file: mageck corrected output
:return: None
"""
pre = pd.read_csv(mo_uncorrected_file, compression='infer', sep="\t", index_col='id')
post = pd.read_csv(mo_corrected_file, compression='infer', sep="\t", index_col='id')
pre.sort_index(inplace=True)
post.sort_index(inplace=True)
pre_genes = set(pre.index.tolist())
preD = set(pre[(pre['neg|fdr'] < fdrth) & (pre['pos|fdr'] >= fdrth)].index.tolist())
preE = set(pre[(pre['neg|fdr'] >= fdrth) & (pre['pos|fdr'] < fdrth)].index.tolist())
preNULL = pre_genes.difference(preD.union(preE))
post_genes = set(post.index.tolist())
postD = set(post[(post['neg|fdr'] < fdrth) & (post['pos|fdr'] >= fdrth)].index.tolist())
postE = set(post[(post['neg|fdr'] >= fdrth) & (post['pos|fdr'] < fdrth)].index.tolist())
postNULL = post_genes.difference(postD.union(postE))
aDD = len(preD.intersection(postD))
aDN = len(preD.intersection(postNULL))
aDE = len(preD.intersection(postE))
aND = len(preNULL.intersection(postD))
aNN = len(preNULL.intersection(postNULL))
aNE = len(preNULL.intersection(postE))
aED = len(preE.intersection(postD))
aEN = len(preE.intersection(postNULL))
aEE = len(preE.intersection(postE))
cm = np.matrix([[aDD, aDN, aDE], [aND, aNN, aNE], [aED, aEN, aEE]])
IMPACTEDg = ((np.triu(cm, 1) + np.tril(cm, -1)) / cm.sum()).sum() * 100
IMPACTED_phenGenes = (cm[0, 1] + cm[2, 1] + cm[0, 2] + cm[1, 2]) / (cm[[0, 2]]).sum() * 100
DISTORTEDg = (cm[2, 0] + cm[0, 2]) / cm.sum() * 100
DISTORTED_phenGenes = (cm[2, 0] + cm[0, 2]) / (cm[[0, 2]]).sum() * 100
cm_fitmat = np.divide(cm.T, np.concatenate([cm.sum(1).T] * 3)) * 100
original_counts = cm.sum(1).T.tolist()[0]
label_list = ['loss_of_fitness', 'no_phenotype', 'gain_of_fitness']
mod_label = ["{}<br>{}</br>".format(original_counts[i], label_list[i]) for i in range(3)]
barplot_list = []
for i in range(3):
trace1 = go.Bar(
x=mod_label,
y=cm_fitmat[i, :].tolist()[0],
name=label_list[i]
)
barplot_list.append(trace1)
layout = go.Layout(
title='Impact on phenotype',
xaxis=dict(
title="Original Counts",
),
yaxis=dict(title='% of genes',
),
barmode='stack',
legend=dict(
x=1,
y=1,
orientation="v"
),
annotations=[
dict(
x=1.17,
y=1.01,
align="right",
valign="top",
text='Corrected Counts',
showarrow=False,
xref="paper",
yref="paper",
xanchor="left",
yanchor="top"
)]
)
figure = go.Figure(data=barplot_list, layout=layout)
py.plot(figure, filename=saveto + '_barchart.html', auto_open=False,
config=PlotData.plotly_conf(cfprm={'theme': 'ggplot'}))
pichart_dict = {('Overall impact', 'Rest of the genes'): [IMPACTEDg, 100 - IMPACTEDg],
('Overall distortion', 'Rest of the genes'): [DISTORTEDg, 100 - DISTORTEDg],
('Impact (G/L fitness genes)', 'Rest of the genes'): [IMPACTED_phenGenes,
100 - IMPACTED_phenGenes],
('Distortion (G/L fitness genes)', 'Rest of the genes'): [DISTORTED_phenGenes,
100 - DISTORTED_phenGenes]}
pi_colors = ['red', 'blue', 'green', 'orange']
pi_x = [[0, .48], [.52, 1], [0, .48], [.52, 1]]
pi_y = [[0, .49], [0, .49], [.51, 1], [.51, 1]]
# pie chart
pie_list = []
count = 0
for label_tuple in pichart_dict:
pie_data = {
'labels': list(label_tuple),
'values': pichart_dict[label_tuple],
'type': 'pie',
'domain': {'x': pi_x[count],
'y': pi_y[count]
},
'marker': {'colors': [pi_colors[count], '#FFFFFF'],
'line': {'color': '#000000',
'width': 0.5},
},
'hoverinfo': 'label+percent',
'textinfo': 'label'
}
count += 1
pie_list.append(pie_data)
layout = go.Layout(
title="Impact of Correction on Genes",
)
figure = go.Figure(data=pie_list, layout=layout)
py.plot(figure, filename=saveto + '_piechart.html', auto_open=False,
config=PlotData.plotly_conf())
return None
def evolution_bars_plot(df):
"""FIXME! briefly describe function
:param df:
:returns:
:rtype:
"""
x = df['Date'].values
fig = go.Figure()
config = {'displayModeBar': True}
fig.add_traces(data=[
go.Bar(name='Décès', x=x, y=df['Décès'].values, marker_color='red'),
go.Bar(name='Guéris',
x=x,
y=df['Guéris'].values,
marker_color='darkgreen'),
go.Bar(name='Cas hospitalisés',
x=x,
y=df['Cas hospitalisés'],
marker_color='gold')
])
fig.add_trace({
'x': df['Date'],
'y': df['Cas confirmés'],
'name': 'Cas confirmés'
})
fig.add_traces(data=[
go.Bar(name='Nouveaux décès',
x=x,
y=df['Nouveaux décès'].values,
marker_color='red',
visible=False),
go.Bar(name='Nouveaux guéris',
x=x,
y=df['Nouveaux guéris'].values,
marker_color='darkgreen',
visible=False),
go.Bar(name='Nouveaux cas',
x=x,
y=df['Nouveaux cas'],
marker_color='gold',
visible=False)
])
fig.update_yaxes(automargin=True)
# Add dropdown
fig.update_layout(
updatemenus=[
dict(type='dropdown',
active=0,
buttons=list([
dict(
label="Cumul",
method="update",
args=[{
"visible":
[True, True, True, True, False, False, False]
}],
),
dict(
label="Par jour",
method="update",
args=[{
"visible":
[False, False, False, False, True, True, True]
}, {
"title":
'Suivi journalier du Covid-19 en Guinée',
}],
)
]),
direction="down",
showactive=True,
x=0.01,
xanchor="center",
y=1.02,
yanchor='middle',
bgcolor='darkred',
font=dict(color='gray', size=14)),
],
template='plotly_dark',
barmode='stack',
hovermode='x',
xaxis_tickangle=-60,
legend_orientation="h",
legend=dict(x=0, y=-0.3),
margin=dict(t=40, b=0, l=25, r=3),
title=dict(text='Évolution du Covid-19 en Guinée : données cumulées',
x=0.5,
y=0.1,
font=dict(size=12.5)),
)
return fig
main_list = []
separator = ','
for item in data.cuisines:
item_list = item.split(',')
new_list = []
for it in item_list:
str_content = it.strip()
new_list.append(str_content)
main_list.append(separator.join(sorted(new_list)))
data['new_cuisines'] = main_list
data['new_cuisines'][0]
# Visualization
location_count = data.groupby(['place']).size().reset_index(name="count")
location_count = location_count.sort_values('count', ascending=True)
location_count.head()
inputdata = [go.Bar(x=location_count['place'], y=location_count['count'])]
py.iplot(inputdata, filename='basic-bar')
plt.show()
rate_group_count = data.groupby(['rate']).size().reset_index(name="count")
rate_group_count = rate_group_count.sort_values('count', ascending=True)
rate_group_count.head()
inputdata = [go.Bar(x=rate_group_count['rate'], y=rate_group_count['count'])]
py.iplot(inputdata, filename='basic-bar')
plt.show()
name_group_count = data.groupby(['name']).size().reset_index(name="count")
name_group_count = name_group_count.sort_values('count', ascending=False)
#name_group_count.head()
top20 = name_group_count[0:20]
top20
inputdata = [go.Bar(x=top20['name'], y=top20['count'])]
py.iplot(inputdata, filename='basic-bar')
import dash_html_components as html
import dash_core_components as dcc
import plotly.graph_objs as go
import numpy as np
import pandas as pd
data = pd.read_csv("/Users/veronikapeskova/projects/portfolio/plotly_dash/apps_app/apps/data/category.csv")
bar_data = [
go.Bar(
y=data['Category'].value_counts().sort_values().to_dense().keys(),
x=data['Category'].value_counts().sort_values(),
orientation='h',
text="Number of Apps in Category",
)]
layout = go.Layout(
height=800,
title='Number of Apps in each category',
hovermode='closest',
yaxis=dict(title='Category', gridwidth=2, domain=[0.1, 1]),
showlegend=False
)
fig = go.Figure(data=bar_data, layout=layout)
graph = html.Div(
[
dcc.Graph(
id='my-graph',
def bar_mul(player_name1: str, player_name2: str, metric1: str, metric2=None):
'''Presents either a singular or grouped bar graph for a pair of players.
Parameters
----------
player_name1: string
The name of the first player being searched for.
player_name2: string
The name of the second player being searched for.
metric1: string
The first metric to be visualized.
metric2 (optional): string
The second metric to be visualized.
Returns
-------
None
'''
f_name1 = player_name1.split()[0]
s_name1 = player_name1.replace(f_name1 + " ", "")
f_name2 = player_name2.split()[0]
s_name2 = player_name2.replace(f_name2 + " ", "")
if metric2:
players = [labels_dict[metric1], labels_dict[metric2]]
fig = go.Figure(data=[
go.Bar(name=player_name1,
x=players,
y=[
new_sample_data.query(
"first_name==@f_name1 and second_name==@s_name1")
[metric1].sum(),
new_sample_data.query(
"first_name==@f_name1 and second_name==@s_name1")
[metric2].sum()
]),
go.Bar(name=player_name2,
x=players,
y=[
new_sample_data.query(
"first_name==@f_name2 and second_name==@s_name2")
[metric1].sum(),
new_sample_data.query(
"first_name==@f_name2 and second_name==@s_name2")
[metric2].sum()
])
])
fig.update_layout(barmode='group')
else:
fig = px.bar(
x=[player_name1, player_name2],
y=[
new_sample_data.query(
"first_name==@f_name1 and second_name==@s_name1")
[metric1].sum(),
new_sample_data.query(
"first_name==@f_name2 and second_name==@s_name2")
[metric1].sum()
],
labels={
'x': labels_dict[metric1],
'y': 'Value'
},
title=f'{player_name1} vs {player_name2}: {labels_dict[metric1]}')
fig.show()
cufflinks.go_offline(connected=True)
init_notebook_mode(connected=True)
plotly.tools.set_credentials_file(username='******',
api_key='d1Ip8twrCBnED72ibDBX')
data = pd.read_csv('antibiotics_data.csv')
posData = data.loc[data["Gram Staining "] == 'positive']
negData = data.loc[data["Gram Staining "] == 'negative']
bacteria = data["Bacteria "]
trace1 = go.Bar(x=bacteria,
y=negData["Penicilin"],
name='Negative Penicilin',
text=negData["Penicilin"],
textposition='auto',
marker=dict(color='rgb(234, 177, 202)', ))
trace2 = go.Bar(x=bacteria,
y=negData["Streptomycin "],
name='Negative Streptomycin',
text=negData["Streptomycin "],
textposition='auto',
marker=dict(color='rgb(243, 208, 208)', ))
trace3 = go.Bar(x=bacteria,
y=negData["Neomycin"],
text=negData["Neomycin"],
name='Negative Neomycin',
textposition='auto',
t =(join_df["Male FTC headcount"]+join_df["Female FTC headcount"]+
join_df["Male OEC headcount"]+join_df["Female OEC headcount"])
join_df["Male FTC Proportion"]=(join_df["Male FTC headcount"]/t)
join_df["Female FTC Proportion"]=(join_df["Female FTC headcount"]/t)
join_df["Male OEC Proportion"]=(join_df["Male OEC headcount"]/t)
join_df["Female OEC Proportion"]=(join_df["Female OEC headcount"]/t)
join_df = join_df.round(decimals=2)
join_df = join_df.drop(["Male FTC headcount","Female FTC headcount","Male OEC headcount","Female OEC headcount"],axis=1)
yaxis=["Male FTC Proportion","Female FTC Proportion","Male OEC Proportion","Female OEC Proportion"]
trace = [go.Bar(
x=[join_df["grade"],join_df["year"]],
y=join_df[i],
name=i
) for i in yaxis]
layout=go.Layout(title=dict(text="Headcount for FTC and OEC by grade by proportion",
y=0.9,
x=0.5,
xanchor="center",
yanchor="top"),
barmode="stack",
yaxis=dict(tickformat="%"),
legend=dict(title=None,
y=1.02, yanchor="bottom",
x=1, xanchor="right",
orientation="h"),
legend_title_text='Click legend to filter data')
def sent_cust(target):
my_path = os.path.abspath(os.path.dirname('__file__'))
path_in_ngrams = os.path.join(my_path, "../data/cpickle/")
path_in_pos = os.path.join(my_path, "../data/pos/")
import colorlover as cl
from IPython.display import HTML
colors_all = cl.scales['11']["qual"]["Set3"]
colors_all.extend(cl.scales['11']["qual"]["Paired"])
colors_all.extend(cl.scales['11']["qual"]["Set3"])
import plotly.plotly as py
import plotly.graph_objs as go
data = []
my_path = os.path.abspath(os.path.dirname('__file__'))
path_in_sent = os.path.join(my_path, "data/yelp_sentiment/")
dat = pickle.load(open(path_in_sent + "yelp_sent.p", "rb"))
dt = {}
mata = []
for aes, dars in dat.items():
i = -1
data = []
dars = (dars / len(dars.columns))/4*100
for col in dars.columns:
i = i + 1
trace = go.Bar(
y=list(dars.index), x=list(dars[col]),
name=str(col),
orientation='h',
legendgroup=str(aes),
marker=dict(color=colors_all[i],
line=dict(color=colors_all[i],
width=1)))
data.append(trace)
mata.append(trace)
layout = go.Layout(
barmode='stack',
bargap=0.2,
title="Service, Food, Preparation and Location Sentiment <br>"
"(Please Use Legend To Toggle)",
#width=500,
height=500,
xaxis=dict(
showgrid=False,
zeroline=False,
showline=False,
),
yaxis=dict(
showgrid=False,
zeroline=False,
showline=False,
),
)
fig = go.Figure(data=data, layout=layout)
# py.iplot(fig, filename='marker-h-bar')
dt[aes] = fig
fig = go.Figure(data=mata, layout=layout)
dt["full"] = fig
### For Quarter:
my_path = os.path.abspath(os.path.dirname('__file__'))
path_in_sent = os.path.join(my_path, "data/yelp_sentiment/")
dat = pickle.load(open(path_in_sent + "yelp_sent.p", "rb"))
ke = []
i = -1
for aes, dars in dat.items():
i = i + 1
ke.append(aes)
if i == 0:
va = pd.DataFrame(index=dars.index)
va = pd.concat((va, dars), axis=1)
ra = va[va.index.str.contains("-TQ")]
df_rank = pd.DataFrame()
for col in ra.columns:
df_rank[col] = ra[col].rank(ascending=0)
df_rank = df_rank.astype(int)
bjri_tq = va[va.index == (target + "-TQ ")]
# .iloc[0,:]
# .sort_values()
bjri_tq = bjri_tq.T
bjri_tq_rank = df_rank[df_rank.index == (target + "-TQ ")]
bjri_tq["rank"] = bjri_tq_rank.T
bjri_tq = bjri_tq.sort_values((target + "-TQ "), ascending=True)
trace0 = go.Bar(
x=bjri_tq[(target + "-TQ ")].values,
y=bjri_tq.index,
orientation='h',
text=["Overall Position: " + str(s) for s in list(bjri_tq["rank"].values)],
marker=dict(
color='rgb(158,202,225)',
line=dict(
color='rgb(8,48,107)',
width=1.5,
)
),
opacity=0.6
)
data = [trace0]
layout = go.Layout(
annotations=[dict(
showarrow=False,
text =str(int(bjri_tq[(target + "-TQ ")].mean()*100)) +'/100',
#xanchor="right",
x=1,
y=3,
xref='x',
yref='y',
opacity=0.1,
font=dict(
color="black",
size=30
),)],
height=800,
xaxis=dict(
showgrid=False,
zeroline=False,
showline=False,
),
yaxis=dict(
showgrid=False,
zeroline=False,
showline=False,
# tickangle=45,
tickfont=dict(
size=10),
),
margin=go.Margin(
l=90,
r=0,
b=0,
t=70,
pad=0
),
title=bjri_tq.columns[0] + ' Sentiment Report',
)
fig = go.Figure(data=data, layout=layout)
dt["quarter"] = fig
ba = va[va.index.str.contains("-TQ")].T.sum()
ba.index = [s[:-4] for s in list(ba.index)]
ba = ba/len(va.columns)*100
trace1 = go.Bar(
x=ba.index,
y=ba.values,
marker=dict(
color='Lightgrey',
line=dict(
color='rgb(8,48,107)',
width=1.5,
)
),
opacity=1
)
layout = go.Layout(
height=150,
xaxis=dict(
showgrid=False,
zeroline=False,
showline=False,
),
yaxis=dict(
showgrid=False,
zeroline=False,
showline=False,),
margin=go.Margin(
l=0,
r=0,
b=23,
t=20,
pad=0
),
)
data = [trace1]
fig_national = go.Figure(data=data, layout=layout)
dt["small"] = fig_national
return dt
def update_histogram(datePicked, selection):
date_picked = dt.strptime(datePicked, "%Y-%m-%d")
monthPicked = date_picked.month - 4
dayPicked = date_picked.day - 1
[xVal, yVal, colorVal] = get_selection(monthPicked, dayPicked, selection)
layout = go.Layout(
bargap=0.01,
bargroupgap=0,
barmode="group",
margin=go.layout.Margin(l=10, r=0, t=0, b=50),
showlegend=False,
plot_bgcolor="#323130",
paper_bgcolor="#323130",
dragmode="select",
font=dict(color="white"),
xaxis=dict(
range=[-0.5, 23.5],
showgrid=False,
nticks=25,
fixedrange=True,
ticksuffix=":00",
),
yaxis=dict(
range=[0, max(yVal) + max(yVal) / 4],
showticklabels=False,
showgrid=False,
fixedrange=True,
rangemode="nonnegative",
zeroline=False,
),
annotations=[
dict(
x=xi,
y=yi,
text=str(yi),
xanchor="center",
yanchor="bottom",
showarrow=False,
font=dict(color="white"),
) for xi, yi in zip(xVal, yVal)
],
)
return go.Figure(
data=[
go.Bar(x=xVal, y=yVal, marker=dict(color=colorVal), hoverinfo="x"),
go.Scatter(
opacity=0,
x=xVal,
y=yVal / 2,
hoverinfo="none",
mode="markers",
marker=dict(color="rgb(66, 134, 244, 0)",
symbol="square",
size=40),
visible=True,
),
],
layout=layout,
)
]
layout = go.Layout(title='Comparison of amazon ratings of Electronic Items')
fig = go.Figure(data=data, layout=layout)
pyo.plot(fig, filename='box.html')
# doing a barplot of average rating and number of ratings for a particular item
item_rating_count = top_items[top_items['Item'] == 'B005CT56F8']
bar_data = item_rating_count.groupby('Month').mean()
bar_data['Counts'] = item_rating_count.groupby('Month').size()
bar_data = bar_data.reset_index()
# ploting bar plot using plotly
months = [
'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct',
'Nov', 'Dec'
]
trace1 = go.Bar(x=months,
y=bar_data['Rating'],
name='Rating',
marker=dict(color='#FFD700'))
trace2 = go.Bar(x=months,
y=bar_data['Counts'],
name='Counts',
marker=dict(color='#9EA0A1'))
data = [trace1, trace2]
layout = go.Layout(
title='Bar diagram- average rating and no_of_counts per each month -2013')
fig = go.Figure(data=data, layout=layout)
pyo.plot(fig, filename='bar.html')
import plotly.graph_objs as go
import plotly.offline as pyo
import pandas as pd
df = pd.read_csv('./salaries.csv')
df = df.sort_values('Salary', ascending=False)
df = df[df['Age'] > 30]
data = [go.Bar(x=df['Name'], y=df['Salary'])]
layout = go.Layout(title='Salary by Name',
xaxis=dict(title='Name'),
yaxis=dict(title='Salary'))
fig = go.Figure(data=data, layout=layout)
pyo.plot(fig)
def plot_InfSites_gens(mrcas,
point_up,
root_lib,
range_theta,
Theta=1,
mut_rate=9e-8,
height=500,
width=900):
from structure_tools.Coal_tools import tree_descent_gen
hap_frame = [[''.join([str(x) for x in z])] for z in mrcas]
hap_frame = list(range(len(mrcas)))
hap_frame = pd.DataFrame(hap_frame, columns=['hap_id'])
hap_frame["hap"] = [''.join([str(x) for x in z]) for z in mrcas]
Ncols = 1
titles = [''.join([str(x) for x in y]) for y in mrcas]
fig = []
vals = []
for gp in range(len(titles)):
title = titles[gp]
sink, starters = get_sinks(mrcas[gp], root_lib, point_up)
t1 = time.time()
if len(starters):
#node_weigths, paths_reverse = tree_descent(root_lib,point_up,sink,init= starters,Theta= thet)
#probe_rec= node_weigths[0][0]
node_weigths, paths_reverse, node_bins, paths_vector = tree_descent_gen(
root_lib, point_up, sink, Theta=Theta, mu=mut_rate)
paths_vector = paths_reverse[0][0]
average_gen = np.mean(paths_vector)
vals.append(average_gen)
sort_vals = np.argsort(vals)
vals = [vals[x] for x in sort_vals]
titles = [titles[x] for x in sort_vals]
fig = [go.Bar(x=['hap: {}'.format(x) for x in range(len(titles))], y=vals)]
layout = go.Layout(title='gens until first hap appearence',
barmode='group',
xaxis=dict(title='hap'),
yaxis=dict(title='Gen'))
Figure = go.Figure(data=fig, layout=layout)
hap_frame['t'] = [round(c, 3) for c in vals]
hap_frame = hap_frame.sort_values('t')
return hap_frame, Figure
print("Graph of OVH")
total_epsilon_per_choice = 0
total_ip_weight = 0
graph_pairs = {}
num_weights = 0
for pair in ovh_data:
ip = pair[0]
weights = pair[1]
total_weight = 0
graph_pairs[ip] = 0
for weight in weights:
total_weight += weight
graph_pairs[ip] += ip_epsilons[ip] * weight
#print(graph_pairs[ip])
# Still need the particular probability that this as picks this guard
total_ip_weight += weight
total_epsilon_per_choice += (ip_epsilons[ip] * weight)
if total_weight > 0:
graph_pairs[ip] = graph_pairs[ip] / total_weight
total_epsilon_per_choice = total_epsilon_per_choice / total_ip_weight
## Graph of epsilon values versus Ip Addresses for Given AS
graph_ips = graph_pairs.keys()
graph_epsilons = []
graph_ips = sorted(graph_ips)
for key in graph_ips:
graph_epsilons.append(graph_pairs[key])
graph_data = [go.Bar(x=graph_ips, y=graph_epsilons)]
py.plot(graph_data, filename='basic-bar.html')
fig.append_trace(trace0, 1, 1)
fig.append_trace(trace1, 1, 1)
fig.append_trace(trace2, 1, 1)
fig.append_trace(trace3, 1, 1)
fig.append_trace(trace4, 1, 1)
number = [age0[-1], age1[-1], age2[-1], age3[-1]]
percent = [round(100.0 * (float(x) / float(total[-1])), 1) for x in number]
age_bracket = ['18-24 yrs', '25-44 yrs', '45-65 yrs', '65+ yrs']
trace5 = go.Bar(
x=age_bracket,
y=percent,
marker=dict(
color='rgba(50, 171, 96, 0.6)',
line=dict(color='rgba(50, 171, 96, 1.0)', width=1),
),
name='Percent distribution by age, July 2017',
)
layout = dict(
title='Total number of homeless adults',
yaxis1=dict(showgrid=False,
showline=False,
showticklabels=True,
domain=[0, 0.85],
range=[0, 5200],
title='Number of homeless adults'),
yaxis2=dict(showgrid=False,
showline=True,
