我已经为时间序列分析建立了一个模型，它将预测未来几天的航行，该模型运行良好，但我想使用anaconda jupyter笔记本将其转换为json格式的rest api，请让我知道方法。提前谢谢。
代码如下：

from pandas import Series
from statsmodels.tsa.arima_model import ARIMA
import numpy

# create a differenced series

def difference(dataset, interval=1):
     diff = list()
for i in range(interval, len(dataset)):
    value = dataset[i] - dataset[i - interval]
    diff.append(value)
return numpy.array(diff)

# invert differenced value

def inverse_difference(history, yhat, interval=1):
   return yhat + history[-interval]

# load dataset

  series = Series.from_csv('mkr.csv', header=None)

# seasonal difference

X = series.values
X = X.astype('float32')
days_in_year = 365
differenced = difference(X, days_in_year)

 # fit model
 model = ARIMA(differenced, order=(0,0,1))
 model_fit = model.fit(disp=0)

# multi-step out-of-sample forecast

 forecast = model_fit.forecast(steps=7)[0]

# invert the differenced forecast to something usable

history = [x for x in X]
day = 1
 for yhat in forecast:
    inverted = inverse_difference(history, yhat,days_in_year)
    print('Day %d sail:= %.3f' % (month, inverted))
    history.append(inverted)
    day += 1