Heatmap Alphabetically Sorting Index Values

Question

Need some help figuring out how to fit this...

df = pd.read_csv('data.csv')
df1 = df.pivot(index='Name', columns='Years', values='Apples')
sns.heatmap(df1)

The data.csv looks like this:

Name	Years	Apples
Henry	2020	2
Bob	2020	2
Johnny	2020	1
Henry	2021	6
Bob	2021	2
Johnny	2021	1

Now when I try to do a heatmap using Seaborn, it prompts out a Value Error stating I can't convert "Henry" from a string to a float.

I figured after googling that I need to pivot the dataframe. Unfortunately, it also alphabetically sorted the "Names" column. Is it possible to maintain the same order as "Henry", "Bob", and "Johnny" instead?

Heatmap Screenshot

Answer 1

If you did a print of df1, you will notice that the sorting is already done... the issue is because of pivot, rather than the plotting itself.

>> print(df1)

Years   2020  2021
Name              
Bob        2     2
Henry      2     6
Johnny     1     1

So, to NOT sort it by index (Name), you can do this using pandas categorical, which will allow you to define the order to keep it in the same order as in the data... So, the updated code (just the one additional line added)...

df.Name=pd.Categorical(df.Name,categories=df.Name.unique(),ordered=True)
df1 = df.pivot(index='Name', columns='Years', values='Apples')
sns.heatmap(df1)

will give you the required plot as below.