如何将csv转换为json

Question

我有以下csv文件:

Topic,Characteristics,Total
Population and dwellings,Population-2016,183314
Population and dwellings,Population-2011,175779
Population and dwellings,Population percentage change,4.3
Age characteristics,0 to 14 years,30670
Age characteristics,0 to 4 years,9275
Age characteristics,5 to 9 years,10475

我想输出一个json文件，这样每个唯一的"主题"是一个键，值是一个"特征":"总"的字典，即输出将是:

{
"Population and dwellings": {
"Population-2016": 183314,
"Population-2011": 175779,
"Population percent change": 4.3
},
"Age characteristics": {
"0 to 14 years": 30670,
"0 to 4 years": 9275,
"5 to 9 years": 10475
}
}

我怎样才能正确地做这件事?到目前为止，我所尝试的所有方法都是相互覆盖的，任何帮助都将不胜感激。谢谢。

Answer 1

可以使用csv模块读取文件，dict.setdefault模块对元素进行分组:

import csv
out = {}
with open("your_file.csv", "r") as f_in:
reader = csv.reader(f_in)
next(reader)  # skip headers
for topic, characteristics, total in reader:
out.setdefault(topic, {})[characteristics] = float(total)
print(out)

打印:

{
"Population and dwellings": {
"Population-2016": 183314.0,
"Population-2011": 175779.0,
"Population percentage change": 4.3,
},
"Age characteristics": {
"0 to 14 years": 30670.0,
"0 to 4 years": 9275.0,
"5 to 9 years": 10475.0,
},
}

---

要从out输出JSON，您可以这样做:

import json
print(json.dumps(out, indent=4))

Answer 2

我的解决方案与Andrej类似，但我使用defaultdict来简化代码。

import collections
import csv
import json

out = collections.defaultdict(lambda: collections.defaultdict(float))
with open("data.csv") as stream:
next(stream)  # Skip the header
reader = csv.reader(stream)
for topic, characteristics, total in reader:
out[topic][characteristics] += float(total)
print(json.dumps(out, indent=4))

输出:

{
"Population and dwellings": {
"Population-2016": 183314.0,
"Population-2011": 175779.0,
"Population percentage change": 4.3
},
"Age characteristics": {
"0 to 14 years": 30670.0,
"0 to 4 years": 9275.0,
"5 to 9 years": 10475.0
}
}