def _update_div1(df):
df_ = pd.read_json(df, orient='split')
fig = px.bar(hastag_counts(df_)[:20], y='#Tags', x='#Num', orientation='h')
fig.update_yaxes(categoryorder='total ascending', title='')
fig.update_layout(margin=dict(t=30, r=0, b=0, pad=1), title_x=.5)
return fig
def top_goalscorer(df):
df['player_rating'] = df['player_name'] + ' - ' + df['rating'].astype(
'str')
fig = px.bar(df.sort_values('goals', ascending=False).head(25),
x='player_rating',
y='goals',
title='Top 25 Goalscorers')
st.plotly_chart(fig)
def draw_interactive_bar_plot_for_simulations(df):
return px.bar(
df,
x='ctp_bin_labels',
y='ctp_bin_counts',
animation_frame='sim_index',
range_y=[0,1300]
)
def create_df(selected_provider):
df1 = psi_data_df[psi_data_df['Provider_Name'] == selected_provider]
results2plot_df = df1.append(benchmark_df, ignore_index=True, sort=True)
return px.bar(results2plot_df,
x='Measure_ID',
y='Rate',
color='Provider_Name',
barmode='group',
hover_name='Measure_Name')
def update_figure(value):
if value == "TE19": df = df_TE19
elif value == "NA19": df = df_NA19
fig = px.bar(df,
y="Närvarograd",
title=f"Närvarograd i procent för {value}")
fig.update_layout(transition_duration=500)
return fig
def draw_interactive_prob_bar_plot_for_simulations(df):
_df = df.copy()
_df['ctp_bin_prob'] = _df['ctp_bin_counts'] / _df['simulations']
return px.bar(_df,
x='ctp_bin_labels',
y='ctp_bin_prob',
animation_frame='sim_index',
range_y=[0, 0.4])
def plot_simple_histogram(self,
metric_name,
x="value",
y="count",
**kwargs):
if not self.has(metric_name):
return
fig = px.bar(self.metrics[metric_name], x=x, y=y, **kwargs)
fig.show()
def top_nations_bar(df):
j = px.bar(df.groupby(['nation_abbrname'
]).size().to_frame('count').sort_values(
['count'],
ascending=False).reset_index().head(25),
x='nation_abbrname',
y='count',
title='Number of Players by Nation')
j.update_xaxes(title='Nation')
j.update_yaxes(title='# of Players')
st.plotly_chart(j)
def top_leagues_bar(df):
i = px.bar(df.groupby(['league_name'
]).size().to_frame('count').sort_values(
['count'],
ascending=False).reset_index().head(25),
x='league_name',
y='count',
title='Number of Players by League')
i.update_xaxes(title='League')
i.update_yaxes(title='# of Players')
st.plotly_chart(i)
def top_clubs_bar(df):
h = px.bar(df.groupby(['club_abbrname'
]).size().to_frame('count').sort_values(
['count'],
ascending=False).reset_index().head(25),
x='club_abbrname',
y='count',
title='Number of Players by Club')
h.update_xaxes(title='Club')
h.update_yaxes(title='# of Players')
st.plotly_chart(h)
def stacked_bar_chart(sample: pd.DataFrame, grouper_col: str, x: str, y: str,
orientation: str, color_col: str, x_title: str,
y_title: str, main_title: str, file: str):
"""
:param sample: A pandas DataFrame
:param grouper_col:
:param x: Column from which x values will be derived
:param y: Column from which y values will be derived
:param orientation: Either 'v' or 'h' for a vertical or horizontal orientation respectively
:param color_col: Column with values to use as colour traces
:param x_title: String for x axis title
:param y_title: String for y axis title
:param main_title: String to input as chart title
:param file: String representing .png filename
:return: A plot.ly graph objects figure
"""
df = sample.copy()
df = df.sort_values(color_col)
pio.templates.default = 'plotly_white'
color_map = color_formatter(sample=df, color_col=color_col)
chart = px.bar(data_frame=df,
x=x,
y=y,
color=color_col,
orientation=orientation,
color_discrete_map=color_map,
title=main_title)
# To ensure that bar heights are arranged in order
df_grouped_for_ordering = df.groupby([grouper_col]).sum().reset_index()
df_grouped_for_ordering = df_grouped_for_ordering.sort_values(
by=[x], ascending=True)
array_ordering = df_grouped_for_ordering['brand'].values
chart.update_layout(barmode='stack',
yaxis={
'categoryorder': 'array',
'categoryarray': array_ordering
})
chart.update_xaxes(title_text=x_title,
showgrid=False,
tickfont=dict(family='Arial', color='Black', size=12),
zeroline=False)
chart.update_yaxes(title_text=y_title,
showgrid=False,
tickfont=dict(family='Arial', color='Black', size=12),
zeroline=False)
chart_formatter(chart=chart, main_title=main_title, legend_pos=-0.15)
pio.write_image(fig=chart, file=filepath + file + '.png')
return chart
def plot_linear_datetimes(self,
time_component,
x="value",
y="count",
**kwargs):
METRIC_NAME = "dates_linear." + time_component
if not self.has(METRIC_NAME):
return
fig = px.bar(self.metrics[METRIC_NAME]["counts"], x=x, y=y, **kwargs)
fig.show()
def bar_chart_2(df):
fig = px.bar(df,
y='Priority',
x="Costs",
color='Campaing',
barmode="group",
orientation="h",
title="POWER BI DASHBOARD")
fig.update_layout(title_x=.5, margin=dict(t=50))
return fig
def bar_chart_2_produc(df):
tmp = df['Products'].value_counts().sort_index().reset_index()
fig = px.bar(tmp,
x='index',
y="Products",
barmode="group",
title="PRODUCTS")
fig.update_layout(title_x=.5, margin=dict(t=50))
return fig
def make_fig2(arg, taxlevel):
levels = {1: "phylum", 2: "class", 3: "order", 4: "family", 5: "genus", 6: "species"}
a = deep[deep["predicted_ARG-class"] == arg]
b = a.groupby(levels[taxlevel]).count()[["predicted_ARG-class"]]
# not use colors if level is species (too many colors)
if taxlevel == 6:
fig = go.Figure(px.bar(b.reset_index(), y="predicted_ARG-class", x=levels[taxlevel], template='plotly_white',
title="Number of " + str(arg) +" resistance genes found per " + str(levels[taxlevel])).for_each_trace(
lambda t: t.update(name=t.name.replace(str(levels[taxlevel]) + "=", ""))))
fig.update_layout(margin=go.layout.Margin(b=200, l=0, r=0))
else:
fig = go.Figure(
px.bar(b.reset_index(), y="predicted_ARG-class", x=levels[taxlevel], template='plotly_white', color=levels[taxlevel],
title="Number of " + str(arg) +" resistance genes found per " + str(levels[taxlevel])).for_each_trace(
lambda t: t.update(name=t.name.replace(str(levels[taxlevel]) + "=", ""))))
fig.update_layout(margin=go.layout.Margin(b=0, l=0, r=0))
fig.update_xaxes(title_text=None)
fig.update_yaxes(title_text="Number of ORFs")
fig.update_layout(autosize=True, titlefont={"size": 20})
fig.update_layout(plot_bgcolor="#F9F9F9", paper_bgcolor="#F9F9F9")
return fig
def analise_dados():
st.header('** Análise das variavéis categoricas**')
st.subheader('Rainfall x Location')
df_process.dropna(inplace=True) # quebra galho retirar
cities_lst = list(df_process['location'].unique())
cities = st.multiselect('Escolha as cidades que deseja ver os dados',
cities_lst,
default=cities_lst)
df_filter = df_process[(df_process['location'].isin(cities))]
soma_local = df_filter.groupby('location')['location',
'rainfall'].sum().reset_index()
soma_local = soma_local.sort_values(by='rainfall')
fig = px.bar(soma_local, x='location', y='rainfall')
st.plotly_chart(fig)
def bar_chart_3_browser(df):
tmp = df.groupby(["Funnel", 'SocialMedia'])['email'].count().reset_index()
fig = px.bar(tmp,
x='SocialMedia',
y="email",
color="Funnel",
barmode="group",
title="REACHABILITY")
fig.update_layout(title_x=.5, margin=dict(t=50))
return fig
def plot_tweets_total(df):
df_count = df.groupby(['handle'], as_index=False).count().iloc[:, 0:2]
df_count.columns = ['handle', 'number of tweets']
# create plot
fig = px.bar(df_count,
x='handle',
y='number of tweets',
color="handle",
color_discrete_map=colour_dict,
title="Number of Tweets",
height=400)
fig.update_layout({"showlegend": False})
fig.update_layout(margin=dict(l=0, r=0, t=30, b=30), autosize=True)
return (fig)
def Exit_Analysis(Employee_ID):
exp = explainer.explain_instance(X.loc[Employee_ID],
predict_fn_ada,
num_features=5)
tmp_lime = pd.DataFrame(exp.as_list(), columns=['feature', 'weight'])
tmp_lime['weight%'] = (round(tmp_lime['weight'] * 10000,
2)).astype(str) + '%'
tmp_lime['Exit'] = (tmp_lime['weight'] > 0).astype('str')
tmp_lime = tmp_lime.replace(regex={'True': 'Yes', 'False': 'No'})
fig = px.bar(
tmp_lime[::-1],
y='feature',
x='weight',
color='Exit',
orientation='h',
color_discrete_map={
'No': 'green',
'Yes': 'red'
},
opacity=0.8,
category_orders={'Exit': ['No', 'Yes']},
text='weight%',
template='plotly_white+presentation+xgridoff',
)
fig = fig.update(layout=dict(title=dict(
text='Top Exit Risk Attributes',
font=dict(family='Arial', size=28, color='black')),
legend=dict(orientation='h', x=0.2, y=1.15),
yaxis=dict(title=None,
ticks='outside',
showline=True,
showgrid=False,
mirror=True,
linecolor='black'),
xaxis=dict(title=None,
showticklabels=False,
showline=True,
mirror=True,
linecolor='black'),
autosize=False,
width=1300,
height=500,
margin=go.layout.Margin(l=350, r=200, t=100)))
return fig
def display_value(value):
if value == "Records lost by Entity":
fig = px.scatter(
df,
x='Records Lost',
y='Year',
range_x=[5000, 3500000000],
symbol='Sector',
animation_group='Sector',
color='Sector',
hover_name='Entity',
hover_data=["Method of Leak"],
color_discrete_sequence=px.colors.qualitative.Alphabet,
log_x=True,
height=900,
opacity=1,
render_mode='webgl',
width=1400)
fig.update_traces(marker=dict(size=15,
line=dict(width=1,
color='DarkSlateGrey')),
selector=dict(mode='markers'))
return fig
elif value == "Records lost by Method":
return go.Figure(data=[
go.Pie(labels=df["Method of Leak"],
values=df["Records Lost"],
hole=0.2)
])
elif value == "Records lost by Year":
return px.bar(df,
x='Year',
y='Records Lost',
color='Method of Leak',
hover_data=['Entity'],
height=800,
width=1400)
elif value == 'Records lost by Sector':
fig_pie = go.Figure(
data=[go.Pie(labels=df["Sector"], values=df["Records Lost"])])
fig_pie.update_traces(hoverinfo='label+percent',
textinfo='value',
textfont_size=10,
marker=dict(colors=df["Sector"],
line=dict(color='#000000',
width=1.2)))
return fig_pie
def redraft_bar(year, df):
d = redraft_data(year, df)
fig = px.bar(d.loc[d.WSPS != 0].sort_values('WSPS', ascending=True),
y='Player',
x='WSPS',
title='Win Shares Per Season (WSPS) by Player',
hover_data=['Pk', 'Redraft', 'Tm', 'College'],
orientation='h')
fig.update_yaxes(tickfont=dict(size=8))
fig.update_xaxes(tickfont=dict(size=8))
fig.update_layout(height=800)
st.plotly_chart(fig)
def get_plot(self, parameters):
target_dict = {
1: "COUNT(*)",
2: "AVG(AvgTone)",
3: "Sum(NumMentions)",
4: "AVG(GoldsteinScale)"
}
qe = QueryExecutor()
query = self.query.format(
month=parameters['month'].value,
role_1=parameters['actor_type'].value,
role_2=(3 - parameters['actor_type'].value),
target_type=target_dict[parameters['target_type'].value],
)
df = qe.get_result_dataframe(query)
df = df.iloc[::-1]
df['Actor1'] = df['Actor1'].map(
Utils().get_fips_country_id_to_name_mapping())
df['Actor2'] = df['Actor2'].map(
Utils().get_fips_country_id_to_name_mapping())
df['Country name'] = df['Actor1'] + ' & ' + df['Actor2']
df['Correlation coefficient'] = df['TargetCorr']
target_name_dict = {
1: 'Event Count',
2: 'Average Tone',
3: 'Sum of Mentions',
4: 'Average Goldstein scale'
}
fig = px.bar(
data_frame=df,
x='Correlation coefficient',
y='Country name',
text='Country name',
orientation='h',
title="Correlation between {} of each day to each country for {}".
format(target_name_dict[parameters['target_type'].value],
Utils().format_months_names([parameters['month'].value
])[0]),
)
fig.update_traces(marker_color='rgb(158,202,225)',
marker_line_color='rgb(8,48,107)',
marker_line_width=1.5,
opacity=0.6)
fig.update_yaxes(title='', showticklabels=False)
return plot(fig, include_plotlyjs=True, output_type='div')
def gen_plotly_datetime(df, args, var):
columns = filter_columns(df, args)
df_temp = df[columns]
df_temp = df_temp[df_temp[args[var]].notnull()]
df_temp = df_temp[['Date', args[var]]]
df_temp.drop_duplicates(inplace=True)
df_temp = pd.DataFrame(df_temp['Date'].value_counts())
df_temp.sort_index(inplace=True)
df_temp.reset_index(inplace=True)
df_temp.columns = ['Date', args.get(var)]
df_temp.rename(columns={args[var]: var}, inplace=True)
fig = px.bar(df_temp, x='Date', y=var, title="{} per day".format(var))
fig.update_xaxes(tickformat="%Y-%m-%d", type="category")
fig.update_layout(showlegend=False)
return fig
def PermImportanceplot(X, y, clf, metric, num_iterations=100):
import plotly_express as px
baseline, scores = PermImportance(X, y, clf, metric, num_iterations)
percent_changes = {c: [] for c in X.columns}
for c in scores:
for i in range(len(scores[c])):
percent_changes[c].append(scores[c][i] / baseline * 100)
return (px.bar(pd.DataFrame.from_dict(percent_changes).melt().groupby(
['variable']).mean().reset_index().sort_values(['value'],
ascending=False)[:25],
x='variable',
y='value',
labels={
'variable': 'column',
'value': '% change in recall'
}))
def gen_plotly_intents(df, threshold=(0.6, 1.0)):
logger.info({"message": "Generating intents plot."})
df_temp = df[[
"response.intents.0.intent", "response.intents.0.confidence"
]]
df_temp.columns = ["intent", "confidence"]
df_temp = df_temp[(df_temp["confidence"] >= threshold[0])
& (df_temp["confidence"] <= threshold[1])]
df_temp = df_temp.groupby("intent").count().sort_values(by="confidence")
df_temp.reset_index(inplace=True)
df_temp.columns = ["intent", "quantity"]
df_temp = df_temp.head(20)
return px.bar(df_temp,
x="quantity",
y="intent",
title="Top 20 intents",
height=600)
def build_Wash_map():
'''Build Simple bar chart of Party choices specific to Washignton State'''
set_pandas()
df_pres = pd.read_csv("../archive/president_county_candidate.csv")
#set party variables
rep = df_pres['party'] == 'REP'
dem = df_pres['party'] == 'DEM'
df_pres = df_pres[rep | dem]
#select just columns where State == Washington
Washginton = df_pres.loc[df_pres['state'] == 'Washington']
#group df based on total votes & aggregate
wash_votes = Washginton.groupby('party')['total_votes'].sum()
fig = px.bar(
wash_votes,
title=
"2020 USA Election: Number of Votes per Party in Washington (Repulcian vs. Democrat)"
)
fig.show()
return
def turnover_rate_plot(feature):
df = data.sort_values(by=[feature, 'Attrition'])
tmp = pd.crosstab(df[feature], df['Attrition']).reset_index()
tmp['sum'] = tmp.sum(axis=1)
tmp['Turnover Rate (%)'] = tmp['Yes'] / (tmp['Yes'] + tmp['No']) * 100
fig = px.bar(
tmp,
x=feature,
y='Turnover Rate (%)',
color=feature,
#text='Turnover Rate (%)',
title='Turnover Rate cross %s' % feature,
template='plotly_white+presentation')
fig = fig.update(layout=dict(showlegend=False,
autosize=False,
width=1100,
height=600,
margin=go.layout.Margin(l=200, r=200, t=100)))
return dcc.Graph(figure=fig)
def update_graph(manager_dropdown_value):
if manager_dropdown_value is None:
return fig_all
else:
df_plot = df.query('Manager == "{}"'.format(manager_dropdown_value))
title = 'Sales funnel of {}'.format(manager_dropdown_value)
orders = (df_plot
.groupby(['Name','Status'])
.Quantity.sum()
.reset_index()
)
fig = px.bar(orders,
x='Name',
y='Quantity',
color='Status',
title=title)
return fig
def fugas_ingresos_por_mes(df):
df_productos = df.loc[:, [
"pk_cid",
"pk_partition",
] + PRODUCTS]
df_productos.drop('pk_cid', axis=1, inplace=True)
clientes_producto = df_productos.groupby('pk_partition').sum()
clientes_mes = pd.DataFrame(clientes_producto.sum(axis=1))
clientes_mes.columns = ['Clientes']
clientes_producto = df_productos.groupby('pk_partition').sum()
clientes_mes['Clientes_Mes_Anterior'] = clientes_mes['Clientes'].shift(1)
clientes_mes['Fugas_Ingresos'] = clientes_mes['Clientes'] - clientes_mes[
'Clientes_Mes_Anterior']
clientes_mes['Fugas_vs_ingresos'] = np.where(
clientes_mes['Fugas_Ingresos'] > 0, 'Ingresos', 'Fugas')
df_fugas_ingresos = clientes_mes[1:]
fig = px.bar(df_fugas_ingresos,
x=df_fugas_ingresos.index,
y='Fugas_Ingresos',
color='Fugas_vs_ingresos',
barmode='group')
return fig
def update_metric_bar_graph(tn, fp, fn, tp, beta):
"""Plot classification metrics as a bar chart"""
m = calc_metrics(tn, fp, fn, tp, beta)
metric_df = pd.DataFrame({
'Metric': list(m.keys()),
'Value': list(m.values())
})
metric_df['display_value'] = round(metric_df['Value'], 3)
fig = px.bar(metric_df,
x='Metric',
y='Value',
color='Metric',
text='display_value')
fig.update_yaxes(range=[0, 1])
fig.layout.update(showlegend=True,
title={
'xanchor': 'center',
'x': 0.5,
'text': 'Accuracy Metrics'
})
return fig
