How to add an extra number on top of the each bar on barchart

Question

According to the explanation why this question is different from this link this link get the height from the diagram as far as I understood, but in my case I do not have this column numpatients6month in the diagram at all, I just have that on the data frame.

So I have a bar chart. It contains two bar for each x-axis in which each bar read from different data frame.

this is the code I am plotting the bar chart.

import seaborn as sns
import matplotlib.pyplot as plt
plt.rcParams['axes.prop_cycle'] = ("cycler('color', 'rg')")
dffinal['CI-noCI']='Cognitive Impairement'
nocidffinal['CI-noCI']='Non Cognitive Impairement'
res=pd.concat([dffinal,nocidffinal])

sns.barplot(x='6month',y='final-formula',data=res,hue='CI-noCI').set_title(fs)
plt.xticks(fontsize=8, rotation=45)

plt.show()

as you see there is two data frame. I plot dffinal with color green and nocidffinal with color red.

This is the result of plot:

Some more explanation: dffinal is based on (6month, final-formula) nocidffinal is also based on(6month,final-formula).

this is my nocidffinal data frame:

          6month  final-formula  numPatients6month
137797.0       1       0.035934                974
267492.0       2       0.021705                645
269542.0       3       0.022107                769
271950.0       4       0.020000                650
276638.0       5       0.015588                834
187719.0       6       0.019461                668
218512.0       7       0.011407                789
199830.0       8       0.008863                677
269469.0       9       0.003807                788
293390.0      10       0.009669                724
254783.0      11       0.012195                738
300974.0      12       0.009695                722

and dffinal:

          6month  final-formula  numPatients6month
166047.0       1       0.077941                680
82972.0        2       0.057208                437
107227.0       3       0.057348                558
111330.0       4       0.048387                434
95591.0        5       0.033708                534
95809.0        6       0.036117                443
98662.0        7       0.035524                563
192668.0       8       0.029979                467
89460.0        9       0.009709                515
192585.0      10       0.021654                508
184325.0      11       0.017274                521
85068.0       12       0.010438                479

As you see there is column numPatients6month in this dataframeS which I would like to show on top of each bar. I do NOT want to change the barchart and group it based on this column, rather I want to just show this number as extra information to the user on top of each bar.

thanks for your time :)

@ImportanceOfBeingErnest can you please unmarked as duplicate? I added explanations how it differs from your link!!! — sariii, Jun 22 '18 at 15:44
I reopened. I think you should also add in how far [this question](https://stackoverflow.com/questions/43214978/seaborn-barplot-displaying-values) does not help. — ImportanceOfBeingErnest, Jun 22 '18 at 16:51
@ImportanceOfBeingErnest sure I will add even more info how they did not help me as I tried three different things I found in stack. Thanks :) — sariii, Jun 22 '18 at 18:50

Josh · Accepted Answer · 2018-06-24T16:36:18.790

If you get your numPatients6month columns in one iterable and in order they appear in chart then using the other stackoverflow answer (also in the docs here) you can place the text on top correctly.

I used code below (adapted from this SO answer). It combines multiple columns one row after another (i.e. will get all your numPatients6month columns in the chart order)

vals = pd.concat([nocidffinal.numPatients6month, dffinal.numPatients6month], axis=1)
vals = vals.stack().reset_index(level=[0,1], drop=True)

This is my full code

import seaborn as sns
import pandas as pd
import matplotlib.pyplot as plt
plt.rcParams['axes.prop_cycle'] = ("cycler('color', 'rg')")
dffinal['CI-noCI']='Cognitive Impairement'
nocidffinal['CI-noCI']='Non Cognitive Impairement'
res=pd.concat([dffinal,nocidffinal])

# Copied to clipboard from SO question above
# Comment out if you already have your dataframes
nocidffinal = pd.read_clipboard().reset_index() 
dffinal = pd.read_clipboard().reset_index() 

# This will merge columns in order of the chart
vals = pd.concat([nocidffinal.numPatients6month, dffinal.numPatients6month], axis=1)
vals = vals.stack().reset_index(level=[0,1], drop=True)

# Plot the chart
ax = sns.barplot(x='6month', y='final-formula', data=res, hue='CI-noCI')
_ = plt.xticks(fontsize=8, rotation=45)

# Add the values on top of each correct bar
for idx, p in enumerate(ax.patches):
    height = p.get_height()
    ax.text(p.get_x()+p.get_width()/2.,
            height + height*.01,
            vals[idx],
            ha="center")

I hope you can help with me to launch the result as you did. I am getting error at nocidffinal = pd.read_clipboard().reset_index() — sariii, Jun 24 '18 at 00:26
Cheers to the response<3 It now works perfectly. I wish I could give double reward :). I am new beginner and firstly I did not realize I should delete this line: pd.read_clipboard().reset_index() maybe its good to update your answer without having those two line for other new beginner like me : — sariii, Jun 24 '18 at 16:30
Yes, glad it works. The `read_clipboard` is for dataframes `nocidffinal` and `dffinal` which I copied to my clipboard from above. For you, you already have these so you can comment those lines out out :) — Josh, Jun 24 '18 at 16:35
yea I figured out how it works and what they are again thanks :) — sariii, Jun 24 '18 at 17:22

How to add an extra number on top of the each bar on barchart

1 Answers1