def plot_original_price_series(df_fut_orig):
"""
Plots Original Corn Futures Price Series using Plotly
"""
fig = tls.make_subplots(rows=5,
cols=1,
shared_xaxes=True,
print_grid=False,
specs=[[{
'rowspan': 4
}], [None], [None], [None], [{}]])
for col in ['Settle']:
fig.append_trace(
{
'x': df_fut_orig.index,
'y': df_fut_orig[col],
'type': 'scatter',
'name': col
}, 1, 1)
for col in ['Volume']:
fig.append_trace(
{
'x': df_fut_orig.index,
'y': df_fut_orig[col],
'type': 'bar',
'name': col
}, 5, 1)
fig['layout']['xaxis'].update(title='Date')
fig['layout']['yaxis1'].update(title='Settling Price (Cents) ')
fig['layout']['yaxis2'].update(title='Volume')
cf.iplot(fig)
def distplot1d(ss: pd.Series,
bins=10,
hist=True,
kde=True,
rug=True,
color=None,
as_figure=False,
legend=True,
title=True,
grid=True,
figsize=None,
**kwargs):
ss = pd.Series(ss)
bin_size = (ss.max() - ss.min()) / bins
curve_type = 'kde' if kde else 'normal'
fig = ff.create_distplot(hist_data=[ss.values],
group_labels=[str(ss.name)],
bin_size=bin_size,
curve_type=curve_type,
colors=color,
show_hist=hist,
show_rug=rug,
**kwargs)
fig['layout'].update(showlegend=legend, title=title)
if figsize:
fig['layout'].update(width=figsize[0], height=figsize[1])
for k, v in fig['layout'].items():
if 'axis' in k:
v.update(showgrid=grid)
if as_figure:
return fig
cf.iplot(fig)
def plot_grouped_by_year_data(df_weekly, title):
"""
Plots weekly combined series (price series and cot report) using Plotly. Use trading weeks on X axis
"""
x_series = np.arange(0, 54)
groups = df_weekly['Settle'].groupby(Grouper(freq='A'))
fig = tls.make_subplots(rows=sum(1 for (name, grp) in groups
if len(grp.values) >= 52),
cols=1,
shared_xaxes=True,
print_grid=False)
fig['layout'].update(height=600, width=899, title=title)
for i, (name, group) in enumerate([(name, grp) for (name, grp) in groups
if len(grp.values) >= 52]):
# chart only data where we have full year (52 weeks)
if len(group.values) >= 52:
fig.append_trace(
{
'x': x_series,
'y': group.values,
'type': 'scatter',
'name': name.year
}, i + 1, 1)
fig['layout']['yaxis' + str((i + 1))].update(showticklabels=False)
cf.iplot(fig)
def plot_fit(series, filename):
y = series.tolist()
x = np.array(range(len(y)))
gauss = Gauss(x, y)
y_pred = gauss.fit()
fit_series = pd.Series(y_pred, series.index, name="Fitted Curve")
_, m, s = gauss.par
current = int(sum(y))
estimate = gauss.estimate_total()
fig1 = series.iplot(kind="bar", asFigure=True)
fig2 = fit_series.iplot(asFigure=True,
colors=['blue'],
width=2,
dash="dashdot")
fig = cf.tools.merge_figures([fig1, fig2])
fig = go.Figure(fig)
fig.update_layout(
title_text="Total Infection Estimate<br>-------------------------------"\
"<br>Current: {} people,"\
" Estimate: {} people".format(current,estimate),
yaxis_title="Capita [-]")
cf.iplot(figure=fig, asUrl=True, filename=filename)
return
def plot_series_to_compare(series1, series2, series1_name, series2_name,
title):
"""
Plots Two series for easy comparison
"""
fig = tls.make_subplots(rows=1,
cols=1,
shared_xaxes=True,
print_grid=False)
fig['layout'].update(height=400, width=899, title=title)
x_series = np.arange(0, len(series1))
fig.append_trace(
{
'x': x_series,
'y': series1,
'type': 'scatter',
'name': series1_name
}, 1, 1)
fig.append_trace(
{
'x': x_series,
'y': series2,
'type': 'scatter',
'name': series2_name
}, 1, 1)
fig['layout']['xaxis'].update(title='Trading Week')
fig['layout']['yaxis'].update(title='Price (Cents)')
cf.iplot(fig)
def distplot(df: pd.DataFrame,
bins=10,
hist=True,
kde=True,
rug=False,
color=None,
as_figure=False,
legend=True,
title=True,
grid=True,
figsize=None,
subplots=False,
layout=None,
sharex=False,
sharey=False,
**kwargs):
"""mimic seaborn.distplot"""
df = pd.DataFrame(df).rename(columns=str)
if not subplots:
hist_data = [v.values for k, v in df.iteritems()]
group_labels = df.columns.tolist()
bin_size = (np.max(df.max()) - np.min(df.min())) / bins
curve_type = 'kde' if kde else 'normal'
fig = ff.create_distplot(hist_data=hist_data,
group_labels=group_labels,
bin_size=bin_size,
curve_type=curve_type,
colors=color,
show_hist=hist,
show_rug=rug,
**kwargs)
else:
figures = [
distplot1d(ss,
bins=bins,
hist=hist,
kde=kde,
rug=rug,
color=color,
as_figure=True,
**kwargs) for _, ss in df.iteritems()
]
fig = tools.get_subplots(figures,
sharex=sharex,
sharey=sharey,
layout=layout)
fig['layout'].update(showlegend=legend, title=title)
if figsize:
fig['layout'].update(width=figsize[0], height=figsize[1])
for k, v in fig['layout'].items():
if 'axis' in k:
v.update(showgrid=grid)
if as_figure:
return fig
cf.iplot(fig)
def returns_tear(self):
"""
Create the following charts:
1. Dollar cumulative return and price vs time
2. Dollar return and price vs time
3. Dollar return distribution
4. View time series
5. Mean return by view
"""
df = self.result
# cum return chart
chart_df = df[['price', 'cum_return']].fillna(method='pad')
chart_df.iplot(title='Cumulative Return and Price',
yTitle='Dollar Cumulative Return',
secondary_y='price',
secondary_y_title='Price')
# return chart
chart_df = df[['price', 'returns']].fillna(method='pad')
chart_df.iplot(title='Return and Price',
yTitle='Dollar Return',
secondary_y='price',
secondary_y_title='Price')
# return distribution chart
chart_df = df.loc[df['view'] != 0, 'returns'].dropna()
chart_df.iplot(title='Return Distribution',
kind='hist',
xTitle='Return',
yTitle='Occurrences')
# view time series
figure = df[['view', 'price']].iplot(title='View and Price',
kind='area',
fill=True,
secondary_y='price',
yTitle='View',
secondary_y_title='Price',
asFigure=True)
figure['data'][0]['line']['shape'] = 'hv'
cf.iplot(figure)
# view statistics
return_group = df.groupby(df.view)['fwd_price_change']
view_stats = return_group.describe()
view_stats['cum_return'] = return_group.sum()
print view_stats
view_stats['mean'].iplot(title='Mean Return by View',
kind='bar',
xTitle='View',
yTitle='Mean Dollar Return')
return
def plot_weekly_combined_series_by_date(df_weekly):
"""
Plots weekly combined series (price series and cot report) using Plotly
"""
fig = tls.make_subplots(rows=5,
cols=1,
shared_xaxes=True,
print_grid=False,
specs=[[{
'rowspan': 2
}], [None], [{
'rowspan': 2
}], [None], [{}]])
for col in ['Settle']:
fig.append_trace(
{
'x': df_weekly['Date'],
'y': df_weekly[col],
'type': 'scatter',
'name': col
}, 1, 1)
for col in ['Open_Interest', 'Longs', 'Shorts']:
fig.append_trace(
{
'x': df_weekly['Date'],
'y': df_weekly[col],
'type': 'scatter',
'name': col
}, 3, 1)
for col in ['Volume']:
fig.append_trace(
{
'x': df_weekly['Date'],
'y': df_weekly[col],
'type': 'bar',
'name': col
}, 5, 1)
fig['layout']['xaxis'].update(title='Date')
fig['layout']['yaxis1'].update(title='Price (Cents)')
fig['layout']['yaxis2'].update(title='Open Interest')
fig['layout']['yaxis3'].update(title='Volume')
cf.iplot(fig)
def plot_cover(df_cover,
df_cover_rand,
label_rand,
title='Percentual de cobertura'):
"""
Plota a curva de precisão-recall para comparação de duas métricas de ordenamento dos alunos (ex: probabilidade de evasão do modelo e idade)
df_cover, df_cover_rand : pandas.DataFrame
Tabelas ordenadas com a métrica de ordenação, precision e recall (gerada pela função calculate_prec_recall)
label_rand, title : str
Nome do modelo de comparação (ex: random, idade) e titulo de grafico, respectivamente
"""
# Plot
data = [
go.Scatter(x=df_cover_rand['perc_cover'],
y=df_cover_rand['precision'],
mode='markers',
marker=dict(color=colorscale[1], opacity=.7),
name='Precision ({})'.format(label_rand)),
go.Scatter(x=df_cover_rand['perc_cover'],
y=df_cover_rand['recall'],
mode='markers',
marker=dict(color=colorscale[5], opacity=.7),
name='Recall ({})'.format(label_rand)),
go.Scatter(x=df_cover['perc_cover'],
y=df_cover['precision'],
mode='lines',
marker=dict(color=colorscale[0]),
name='Precision'),
go.Scatter(x=df_cover['perc_cover'],
y=df_cover['recall'],
mode='lines',
marker=dict(color=colorscale[4]),
name='Recall')
]
layout = dict(xaxis=dict(
title='% de alunos cobertos (ordenado pela prob. de evasão)'),
title=title)
fig = go.Figure(data=data, layout=layout)
cf.iplot(fig)
def plot_roc(y_test, y_score, title, colors=colorscale):
"""
Plota a curva ROC para os dados de treino e teste (preditos).
y_test : numpy.array
y_score : pandas.DataFrame (scores dos modelos a serem comparados)
title : str
"""
ref = np.linspace(0, 1)
data = [
go.Scatter(x=ref,
y=ref,
mode='markers',
marker=dict(color='grey'),
showlegend=False)
]
i = 0
for model in y_score.columns:
# Compute ROC curve and ROC area for each class
fpr, tpr, _ = roc_curve(y_test, y_score[model], pos_label=1)
roc_auc = auc(fpr, tpr)
data.append(
go.Scatter(x=fpr,
y=tpr,
mode='lines',
marker=dict(color=colors[i]),
name='{} (area = {})'.format(model, round(roc_auc, 2))))
i += 1
layout = dict(xaxis=dict(title='Taxa de falsos positivos (TFP)'),
yaxis=dict(title='Taxa de verdadeiros positivos (TVP)'),
title=title)
fig = go.Figure(data=data, layout=layout)
cf.iplot(fig)
def mplot(df: pd.DataFrame,
kind='scatter',
subplots=False,
sharex=False,
sharey=False,
layout=(1, -1),
figsize=None,
title=None,
xlabel=True,
ylabel=True,
legend=True,
xlim=None,
ylim=None,
as_figure=False,
grid=True,
**kwargs):
name = df.columns.name if isinstance(df, pd.DataFrame) else df.name
title = title or name
xTitle = xlabel
if isinstance(xlabel, bool) and xlabel:
xTitle = df.index.name
yTitle = ylabel
if isinstance(ylabel, bool) and ylabel:
yTitle = name
dist_kws = kwargs.pop('dist_kws', dict())
if kind == 'dist':
fig = distplot(df,
as_figure=True,
subplots=subplots,
sharex=sharex,
sharey=sharey,
legend=legend,
title=title,
layout=layout,
grid=grid,
figsize=figsize,
**dist_kws)
# fig['layout'].update(width=figsize[0], height=figsize[1])
else:
fig = df.iplot(kind=kind,
subplots=subplots,
shared_xaxes=sharex,
shared_yaxes=sharey,
showlegend=legend,
dimensions=figsize,
title=title,
xTitle=xTitle,
yTitle=yTitle,
shape=layout,
asFigure=True,
xrange=xlim,
yrange=ylim,
**kwargs)
layout = fig.setdefault('layout', dict())
layout.update(showlegend=legend)
for k, v in layout.items():
if k.startswith(('xaxis', 'yaxis')):
v.update(showgrid=grid)
if True:
if k.startswith('xaxis'):
v.update(title=xTitle)
if xlim:
v.update(range=xlim)
if k.startswith('yaxis'):
v.update(title=yTitle)
if ylim:
v.update(range=ylim)
if as_figure:
return fig
cf.iplot(fig)
def pairplot(df: pd.DataFrame,
kind='scatter',
diag_kind='dist',
theme=None,
bins=10,
color='grey',
size=2,
figsize=None,
title=False,
as_figure=False,
sharex=False,
sharey=False,
grid=False):
""""""
if not theme:
theme = cf.auth.get_config_file()['theme']
figs = []
for coly in df.columns:
for colx in df.columns:
if colx == coly:
if diag_kind == 'dist':
fig = distplot(df[colx],
hist=False,
as_figure=True,
rug=False,
bins=bins)
else:
fig = df.iplot(kind=diag_kind,
keys=[colx],
asFigure=True,
bins=bins)
figs.append(fig)
else:
figs.append(
df.iplot(kind=kind,
mode='markers',
x=colx,
y=coly,
asFigure=True,
size=size,
colors=[color]))
layout = tools.get_layout(theme=theme)
layout['xaxis1'].update(showgrid=grid)
layout['yaxis1'].update(showgrid=grid)
fig = tools.get_subplots(figs,
layout=(len(df.columns), len(df.columns)),
sharex=sharex,
sharey=sharey,
horizontal_spacing=.05,
vertical_spacing=.07,
base_layout=layout)
if isinstance(title, bool) and title:
fig['layout'].update(title=df.columns.name)
elif title:
fig['layout'].update(title=title)
if figsize:
fig['layout'].update(width=figsize[0], height=figsize[1])
fig['layout'].update(bargap=.02, showlegend=False)
if as_figure:
return fig
cf.iplot(fig)
df.iplot(subplots=True, subplot_titles=True, legend=False)
df = cf.datagen.bubble(10, 50, mode='stocks')
figs = cf.figures(df, [
dict(kind='histogram', keys='x', color='blue'),
dict(kind='scatter', mode='markers', x='x', y='y', size=5),
dict(kind='scatter', mode='markers', x='x', y='y', size=5, color='teal')
],
asList=True)
figs.append(
cf.datagen.lines(1).figure(bestfit=True,
colors=['blue'],
bestfit_colors=['pink']))
base_layout = cf.tools.get_base_layout(figs)
sp = cf.subplots(
figs,
shape=(3, 2),
base_layout=base_layout,
vertical_spacing=.15,
horizontal_spacing=.03,
specs=[[{
'rowspan': 2
}, {}], [None, {}], [{
'colspan': 2
}, None]],
subplot_titles=['Histogram', 'Scatter 1', 'Scatter 2', 'Bestfit Line'])
sp['layout'].update(showlegend=False)
cf.iplot(sp)
help(df.iplot)
autosize=True,
showlegend=True,
legend=dict(
font=dict(
family='sans-serif',
#size=20,
color='#000'), ),
title=dict(
text="Global Fatalities",
font=(
dict( #size=40,
color='#000'))),
# Add annotations in the center of the donut pies.
annotations=[dict(text=total_deaths, align="center", showarrow=False)])
cf.iplot(figure=fig, filename=plot_folder + "/death", asUrl=True)
def plot_fit(series, filename):
y = series.tolist()
x = np.array(range(len(y)))
gauss = Gauss(x, y)
y_pred = gauss.fit()
fit_series = pd.Series(y_pred, series.index, name="Fitted Curve")
_, m, s = gauss.par
current = int(sum(y))
estimate = gauss.estimate_total()
fig1 = series.iplot(kind="bar", asFigure=True)
fig2 = fit_series.iplot(asFigure=True,