In [1]:

    

from bokeh.io import output_notebook, show
from bokeh.plotting import figure
from bokeh.palettes import Blues4
from bokeh.transform import factor_cmap
import bokeh.models as bm
import pandas as pd
import chartify
output_notebook()


    

    






    

        
        Loading BokehJS ...
    






    

In [2]:

    

df = pd.read_csv('tmp_df2_percentual_and_absolute.csv', index_col=0)
df.head()


    

    
Out[2]:







  

    

      

      
similarity
      
group
      
absolute
      
percentual
    
  
  

    

      
0
      
100%
      
script_url
      
287460
      
0.348066
    
    

      
1
      
100%
      
script_url_clean
      
363694
      
0.440373
    
    

      
2
      
100%
      
script_url_plus_1
      
656672
      
0.795120
    
    

      
3
      
80%+
      
script_url
      
90798
      
0.109941
    
    

      
4
      
80%+
      
script_url_clean
      
140444
      
0.170054
    
  

In [3]:

    

df.head()


    

    
Out[3]:







  

    

      

      
similarity
      
group
      
absolute
      
percentual
    
  
  

    

      
0
      
100%
      
script_url
      
287460
      
0.348066
    
    

      
1
      
100%
      
script_url_clean
      
363694
      
0.440373
    
    

      
2
      
100%
      
script_url_plus_1
      
656672
      
0.795120
    
    

      
3
      
80%+
      
script_url
      
90798
      
0.109941
    
    

      
4
      
80%+
      
script_url_clean
      
140444
      
0.170054
    
  

In [4]:

    

lookup = {'100%': 0, '80%+': 1, '50%+': 2, '50%-': 3}
df['similarity_index'] = df.similarity.apply(lambda x: lookup[x])
df = df.sort_values(by=['group', 'similarity_index'])
df['top'] = df.groupby('group').cumsum()['percentual']
df['bottom'] = df.top - df.percentual
df['absolute_text'] = df.absolute.apply(lambda x: f'n = {x:,}')
df


    

    
Out[4]:







  

    

      

      
similarity
      
group
      
absolute
      
percentual
      
similarity_index
      
top
      
bottom
      
absolute_text
    
  
  

    

      
0
      
100%
      
script_url
      
287460
      
0.348066
      
0
      
0.348066
      
0.000000
      
n = 287,460
    
    

      
3
      
80%+
      
script_url
      
90798
      
0.109941
      
1
      
0.458007
      
0.348066
      
n = 90,798
    
    

      
6
      
50%+
      
script_url
      
311461
      
0.377127
      
2
      
0.835134
      
0.458007
      
n = 311,461
    
    

      
9
      
50%-
      
script_url
      
136159
      
0.164866
      
3
      
1.000000
      
0.835134
      
n = 136,159
    
    

      
1
      
100%
      
script_url_clean
      
363694
      
0.440373
      
0
      
0.440373
      
0.000000
      
n = 363,694
    
    

      
4
      
80%+
      
script_url_clean
      
140444
      
0.170054
      
1
      
0.610427
      
0.440373
      
n = 140,444
    
    

      
7
      
50%+
      
script_url_clean
      
241902
      
0.292903
      
2
      
0.903330
      
0.610427
      
n = 241,902
    
    

      
10
      
50%-
      
script_url_clean
      
79838
      
0.096670
      
3
      
1.000000
      
0.903330
      
n = 79,838
    
    

      
2
      
100%
      
script_url_plus_1
      
656672
      
0.795120
      
0
      
0.795120
      
0.000000
      
n = 656,672
    
    

      
5
      
80%+
      
script_url_plus_1
      
81609
      
0.098815
      
1
      
0.893935
      
0.795120
      
n = 81,609
    
    

      
8
      
50%+
      
script_url_plus_1
      
64849
      
0.078521
      
2
      
0.972456
      
0.893935
      
n = 64,849
    
    

      
11
      
50%-
      
script_url_plus_1
      
22748
      
0.027544
      
3
      
1.000000
      
0.972456
      
n = 22,748
    
  

In [5]:

    

source = bm.ColumnDataSource(df)
similarities = list(df.similarity.unique())
groups = list(df.group.unique())
p = figure(x_range=groups, y_range=(0, 1), tools='')
p.vbar(
    x='group', 
    bottom='bottom',
    top='top', 
    width=0.9,
    color=factor_cmap('similarity', palette=Blues4, factors=similarities),
    source=source,
    legend='similarity',
)
text_palette = [Blues4[-1]] * 2 + [Blues4[0]] * 2
p.text(
    x='group',
    y='top',
    text='absolute_text',
    source=source,
    text_color=factor_cmap('similarity', palette=text_palette, factors=similarities),
    text_align='center',
    text_font_size='8pt',
    y_offset=15
    
)
p.yaxis.minor_tick_out = None
p.legend.location = (p.plot_width - 150, 10)
#p.yaxis.axis_label = "%"
p.yaxis.formatter = bm.NumeralTickFormatter(format='1%')
#p.add_tools(bm.HoverTool(tooltips="<p>Absolute: @absolute{,}</p><p>Similarity: @similarity</p>"))
p.toolbar_location = None
show(p)


    

    




BokehDeprecationWarning: 'legend' keyword is deprecated, use explicit 'legend_label', 'legend_field', or 'legend_group' keywords instead






    











  








    

In [6]:

    

p = chartify.Chart(x_axis_type='categorical', blank_labels=True)
p.style.set_color_palette('sequential')
p.set_title("Chartified")
p.set_subtitle("Some additional stuff")
p.plot.bar_stacked(
    data_frame=df,
    categorical_columns=['group'],
    numeric_column='percentual',
    stack_column='similarity',
    stack_order=similarities[::-1],
)
p.figure.tools = []
p.figure.y_range.end = 1.2
p.figure.yaxis.minor_tick_out = None
p.axes.set_yaxis_tick_format('1%')
p.show()


    

    




---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-6-b336124d0124> in <module>
----> 1 p = chartify.Chart(x_axis_type='categorical', blank_labels=True)
      2 p.style.set_color_palette('sequential')
      3 p.set_title("Chartified")
      4 p.set_subtitle("Some additional stuff")
      5 p.plot.bar_stacked(

AttributeError: module 'chartify' has no attribute 'Chart'

In [ ]: