如何在 Python 中使用两个单独的尺度将两个数据集聚类为一个热图？

Question

I am trying to perform clustering analysis on two datasets using the cluster heatmap function in Seaborn.

The problem is that the two datasets come from two different procedures and thus they contain values that are differently distributed (I mean, the 1st dataset has a scale of values ranging from 0 to 1, but the 2nd from 1000 up to 5000).

My question is:

How can I cluster two datasets that have different range of values? Is there a way to cluster the rows of the datasets into a single heatmap and maybe have two scales for each dataset?

Here what I have tried so far, but with only little success:

#First, I have combined the two datasets into one dataframe object:
dataset = pd.concat([dataset_1, dataset_2], axis=0)

#Then, passed the dataframe into Seaborn's `.clustermap()` function:
sns.clustermap(data=dataset, 
    col_cluster=False)

Output: you can notice that the features of dataset_1 are all blocked out because of the scale of the difference in scale between the datasets ( dataset_1 and dataset_2 as shown bellow)

Any idea how to approach this problem?

Answer 1

You could use sklearn's preprocessing library, specifically the scaler before creating the clustermap.

The documentation is here: http://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.scale.html#sklearn.preprocessing.scale