Matplotlib grouped data clustered bar chart in Python

Question

I have a dictionary of values (drug) as follows:

{0: {0: 100.0, 1: 0.41249706379061035, 2: 5.144449764434768, 3: 31.078456871927678}, 1: {0: 100.0, 1: 0.6688801420346955, 2: 77.32360971119694, 3: 78.15132480853421}, 2: {0: 100.0, 1: 136.01949766418852, 2: 163.4967732211563, 3: 146.7726208999281}}

It contains 3 drug types, then the efficacy of that drug type in 4 different concentrations.

I'm trying to make a clustered bar chart which compares the 3 drugs against each other like this:

Currently, my code is the following:

fig, ax = plt.subplots()
width = 0.35
ind = np.arange(3)

for x in range(3):
    ax.bar(ind + (width * x), drug[x].values(), width, bottom=0)

ax.set_title('Drug efficacy')
ax.set_xticks(ind + width / 2)
ax.set_xticklabels(list(string.ascii_uppercase[0:drugCount]))

ax.autoscale_view()

plt.show()

I have adapted the code from this guide, but am having multiple problems.

I think the main reason is that the data used in the example is such that values in one group correspond to the same colour rather than the same cluster.

How can I adapt this code such that it will plot the efficacy of each drug in the 4 different concentrations in isolation compared to the other drugs?

Try pd.DataFrame.from_dict(drug_dict, orient='index').plot.bar()? — Zero
– Zero, Commented Oct 12, 2020 at 18:36

Henry Yik · Accepted Answer · 2020-10-12 18:40:24Z

1

IIUC you want to normalize your values by column, which can be done using sklearn:

from sklearn import preprocessing

df = pd.DataFrame(drug)
scaler = preprocessing.MinMaxScaler()
df = pd.DataFrame(scaler.fit_transform(df))
df.T.plot(kind="bar")
plt.show()

answered Oct 12, 2020 at 18:40

Henry Yik

22.6k5 gold badges21 silver badges44 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

Tom Over a year ago

Thanks for your reply, I'll give that a go - any idea why 0 does not show as 100 in this situation yet 2 does for drug 2?

Henry Yik Over a year ago

Not sure If i get it correctly, but when you normalize the values to 0-1 the lowest value becomes 0. Thus in drug 2 the first one becomes 0.

Collectives™ on Stack Overflow

Matplotlib grouped data clustered bar chart in Python

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related