将 Pandas DataFrame 保存到 Django 模型
声明：本页面是StackOverFlow热门问题的中英对照翻译，遵循CC BY-SA 4.0协议，如果您需要使用它，必须同样遵循CC BY-SA许可，注明原文地址和作者信息，同时你必须将它归于原作者（不是我）：StackOverFlow
原文地址: http://stackoverflow.com/questions/37688054/
Warning: these are provided under cc-by-sa 4.0 license. You are free to use/share it, But you must attribute it to the original authors (not me):
StackOverFlow
Saving a Pandas DataFrame to a Django Model
提问by Johan
I have stock price data that is stored in a pandas DataFrame as shown below (actually it was in a panel, but I converted it to a DataFrame)
我的股票价格数据存储在 Pandas DataFrame 中，如下所示（实际上它在面板中，但我将其转换为 DataFrame）
date ticker close tsr
0 2013-03-28 abc 22.81 1.000439
1 2013-03-28 def 94.21 1.006947
2 2013-03-28 ghi 95.84 1.014180
3 2013-03-28 jkl 31.80 1.000000
4 2013-03-28 mno 32.10 1.003125
...many more rows
I want to save this in a Django model, which looks like this (matches the column names):
我想将它保存在一个 Django 模型中，它看起来像这样（匹配列名）：
class HistoricalPrices(models.Model):
ticker = models.CharField(max_length=10)
date = models.DateField()
tsr = models.DecimalField()
close = models.DecimalField()
The best I've come up so far is using this to save it, where df is my DataFrame:
到目前为止，我想到的最好的方法是使用它来保存它，其中 df 是我的 DataFrame：
entries = []
for e in df.T.to_dict().values():
entries.append(HistoricalPrices(**e))
HistoricalPrices.objects.bulk_create(entries)
Is there a better way to save this?
有没有更好的方法来保存它？
I've looked at django-pandas, but looks like it just reads from the DB.
我看过django-pandas，但看起来它只是从数据库中读取。
回答by Stefan
It would be most efficient to use to_sql()with appropriate connectionparameters for the engine, and run this inside your Djangoapp rather than iterating through the DataFrameand saving one modelinstance at a time:
使用to_sql(), 的适当connection参数engine并在您的Django应用程序中运行它是最有效的，而不是一次迭代DataFrame并保存一个model实例：
from django.conf import settings
user = settings.DATABASES['default']['USER']
password = settings.DATABASES['default']['PASSWORD']
database_name = settings.DATABASES['default']['NAME']
database_url = 'postgresql://{user}:{password}@localhost:5432/{database_name}'.format(
user=user,
password=password,
database_name=database_name,
)
engine = create_engine(database_url, echo=False)
df.to_sql(HistoricalPrices, con=engine)
