Here is an alternate visualization of a funnel chart - instead of using bars, it depicts the different steps as concentric circles. How to use: define an array in the extract() function of the values you'd like to use. If you are pulling data from a data frame, you will need to use something like the pandas iloc function to extract the desired values. import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
import matplotlib.patches as mpatches
from matplotlib.collections import PatchCollection
# extract important variables from the data frame
def extract():
# GET DATA FROM DATA FRAME
# if df.size == 0:
# return [0, 0, 0 , 0, 0]
# else:
# opps_created = df.iloc[0]['opps_created']
# opps_accepted = df.iloc[0]['opps_accepted']
# opps_trialed = df.iloc[0]['opps_trialed']
# opps_won = df.iloc[0]['opps_won']
# EXAMPLE DATA
opps_created = 150
opps_accepted = 80
opps_trialed = 50
opps_won = 20
return [opps_created, opps_accepted, opps_trialed, opps_won]
####################
# create text chart
####################
def kpi_text(ax, text = '', text_color = 'black', x_position = 0, y_position = 0.5, font_size = 13):
return ax.text(x = x_position,
y = y_position,
s = text,
color = text_color,
family = 'sans-serif',
fontsize = font_size,
horizontalalignment = 'center',
verticalalignment = 'center')
# create the plot
def kpi_chart(df):
# set up the figure
fig, ax = plt.subplots(figsize = (5.5,5.5))
fig.subplots_adjust(left=0, bottom=0, right=1, top=1, wspace=0, hspace=0)
ax.axis('off')
# styling (colors and font sizes)
mediumgray = '#999292'
lightgray = '#c4c4c4'
purples = plt.get_cmap('Purples_r')
fontsize_text = 30
offset_x = 0.1
offset_y = 0.13
# parse the data
funnel_df = extract()
# use the following if you are using the extract_df function to parse the desired values
# funnel_df = extract(df)
[opps_created, opps_accepted, opps_trialed, opps_won] = funnel_df
# map the data
def get_data_domain(data):
return {'min': min(data), 'max': max(data)}
def data_mapper(data = [None], outputsize = 0.5):
#assumes zero-max domain normalization
# 0.48 is the max radius with a bit of padding
domain = get_data_domain(data)
mapped_area = [1.0 * d/domain['max'] * outputsize for d in data]
mapped_radius = [np.sqrt(d/np.pi) for d in mapped_area]
return mapped_radius
# mapping the data
mapped_df = data_mapper(funnel_df)
colors = ['aliceblue', 'paleturquoise', 'darkturquoise', 'lightseagreen']
# the circles
for i in range(len(mapped_df)):
ax.add_patch(
mpatches.Circle( (0.5 + offset_x, mapped_df[i] + offset_y),
mapped_df[i],
color = colors[i]))
kpi_text(ax,
text = funnel_df[-1],
text_color = 'white',
font_size = fontsize_text,
x_position = 0.5 + offset_x,
y_position = mapped_df[-1] + 0.01 + offset_y)
for i in range(len(mapped_df) - 2, -1, -1):
kpi_text(ax,
text = funnel_df[i],
text_color = 'teal',
font_size = fontsize_text,
x_position = 0.5 + offset_x,
y_position = mapped_df[i + 1] + mapped_df[i] + 0.01 + + offset_y)
cols = ['Created', 'Accepted', 'Trialed', 'Won']
for i in range(len(cols) - 2, -1, -1):
kpi_text(ax,
text = cols[i],
text_color = 'steelblue',
font_size = fontsize_text/1.7,
x_position = 0.5 + offset_x,
y_position = mapped_df[i] + mapped_df[i+1] - 0.05 + 0.01 + offset_y)
kpi_text(ax,
text = cols[-1],
text_color = 'white',
font_size = fontsize_text/1.7,
x_position = 0.5 + offset_x,
y_position = mapped_df[-1] - 0.05 + 0.01 + offset_y)
return plt
periscope.image(kpi_chart())
