从 python 中的嵌套字典列表创建热图

Question

I'm pretty new to python, and currently I'm stuck with writing a heatmap for comparing word vectors for different languages.

The data I have looks like this:

data = [{'de':  {'de': [[1]], 'es': [[0.9644323]], 'fr': [[0.9419257]], 'hu': [[0.9297902]]}},
        {'fr':   {'fr': [[1]], 'de': [[0.9419257]], 'es': [[0.9382719]], 'hu': [[0.91247433]]}},
        {'hu':  {'hu': [[1]], 'de': [[0.9419257]], 'es': [[0.9382719]], 'fr': [[0.91247433]]}}]

The data shows similarity between the phrase in one language to its translations (eg in the first dictionary I have a German phrase that I compare to ES, HU and FR translations). I would like to have one heatmap for this list of nested dictionaries. How do I visualise it in a heatmap? I tried looking at heatmaps in seaborn, but was confused by the fact my dictionary is nested.

Answer 1

If solving quickly. You first transform your list with dictionaries to dataframe and then plot it with seaborn's heatmap. I am sure there could be more elegant solutions, however.

import pandas as pd
import seaborn as sns
%matplotlib inline

frames = []

for d in data:
    frames.append(pd.DataFrame.from_dict(d, orient='index'))

res = pd.concat(frames)
res = res.applymap(lambda x: x[0][0])
sns.heatmap(res, annot=True)