"Python\/Pandas\/Colab - 如何将 numpy.ndarray 转换为 JSON？"

Question

我有一个dbf<\/code>文件，我想使用 Google Colab 将其转换为 JSON 文件。

为了完成这项任务，我将 dbf 加载到pandas dataframe<\/code>框中，然后从那里转换为numpy.ndarray<\/code> ，然后转换为最终的 JSON。

这些是来自 DBF 文件的pd dataframe<\/code>的前 5 行：

打印 (df)

       CD_MUN                 NM_MUN SIGLA_UF  AREA_KM2
0     1100015  Alta Floresta D'Oeste       RO  7067.127
1     1100023              Ariquemes       RO  4426.571
2     1100031                 Cabixi       RO  1314.352
3     1100049                 Cacoal       RO  3793.000
4     1100056             Cerejeiras       RO  2783.300

Answer 1

我确实找到了方法：

import pandas as pd
from simpledbf import Dbf5
import json
import numpy


dbfPath = 'BR_Municipios_2020.dbf'

# Create a 'pandas.core.frame.DataFrame' from the DBF
df = Dbf5(dbfPath).to_dataframe()
# print (type(df)) # <class 'pandas.core.frame.DataFrame'>
# print (df)
# print ('-----') 

# Get an array of values from the 'pandas.core.frame.DataFrame' object
dfValues = df.values
# print (type(dfValues)) #<class 'numpy.ndarray'>
# print (dfValues)
# print ('-----') 

# Convert the 'numpy.ndarray' into a <class 'list'>
lists = dfValues.tolist()
# print (type(lists)) # <class 'list'>
# print (lists)
# print ('-----')

# Traverse the <class 'list'> and recreate a new one
GIS_list = [{'CD_MUN': x[0], 'NM_MUN': x[1], 'SIGLA_UF': x[2]} for i, x in enumerate(lists)]
# print (type(GIS_list)) # <class 'list'>
# print(GIS_list) # [{'CD_MUN': '1100015', 'NM_MUN': "Alta Floresta D'Oeste", 'SIGLA_UF': 'RO'}, {'CD_MUN': '1100023',
# print ('-----')

# Convert the GIS_list into GIS_br_JSON
GIS_br_JSON = json.dumps(GIS_list)
# print (type(GIS_br_JSON)) # <class 'str'>
# print (GIS_br_JSON)
# print ('-----')

# Save the JSON in a file
with open('BR_Municipios_2020.json', 'a') as testwritefile:
    testwritefile.write(GIS_br_JSON)
print ('JSON exported')


# Faster way after Quang Hoang comments
dfDic = df.to_dict(orient='records')
# print (type(dfDic)) # <class 'list'>
# print (dfDic)
# print ('-----') 

GIS_br_JSON2 = json.dumps(dfDic)
print (type(GIS_br_JSON2)) # <class 'str'>
print (GIS_br_JSON2)
print ('-----')

# Save the JSON in a file
with open('BR_Municipios_2020.json', 'a') as testwritefile:
    testwritefile.write(GIS_br_JSON2)
print ('JSON exported')

这是 JSON 验证器中的输出：

@Quang Hoang 评论后更快的方式：

现在的另一个问题是修复这些奇怪的字符： Guajar\á-Mirim