问题描述
我有每月数据，并使用一些额外的回归变量将y转换为log（y）以使用Uncaught TypeError: Cannot read property '_handle' of undefined
at :3000/static/js/0.chunk.js:168900
at Array.forEach (<anonymous>)
at push../node_modules/set-blocking/index.js.module.exports (:3000/static/js/0.chunk.js:168897)
at Object.<anonymous> (:3000/static/js/0.chunk.js:126738)
at Object../node_modules/npmlog/log.js (:3000/static/js/0.chunk.js:127102)
at __webpack_require__ (:3000/static/js/bundle.js:785)
at fn (:3000/static/js/bundle.js:151)
at Object.<anonymous> (:3000/static/js/0.chunk.js:127799)
at Object../node_modules/opencv-build/build/utils.js (:3000/static/js/0.chunk.js:128015)
at __webpack_require__ (:3000/static/js/bundle.js:785)
at fn (:3000/static/js/bundle.js:151)
at Object.<anonymous> (:3000/static/js/0.chunk.js:127280)
at Object../node_modules/opencv-build/build/dirs.js (:3000/static/js/0.chunk.js:127306)
at __webpack_require__ (:3000/static/js/bundle.js:785)
at fn (:3000/static/js/bundle.js:151)
at Object../node_modules/opencv-build/build/constants.js (:3000/static/js/0.chunk.js:127242)
at __webpack_require__ (:3000/static/js/bundle.js:785)
at fn (:3000/static/js/bundle.js:151)
at Object../node_modules/opencv-build/build/index.js (:3000/static/js/0.chunk.js:127599)
at __webpack_require__ (:3000/static/js/bundle.js:785)
at fn (:3000/static/js/bundle.js:151)
at Object.<anonymous> (:3000/static/js/0.chunk.js:128064)
at Object../node_modules/opencv4nodejs/lib/cv.js (:3000/static/js/0.chunk.js:128136)
at __webpack_require__ (:3000/static/js/bundle.js:785)
at fn (:3000/static/js/bundle.js:151)
at Object.<anonymous> (:3000/static/js/0.chunk.js:128154)
at Object../node_modules/opencv4nodejs/lib/opencv4nodejs.js (:3000/static/js/0.chunk.js:128159)
at __webpack_require__ (:3000/static/js/bundle.js:785)
at fn (:3000/static/js/bundle.js:151)
at Module.<anonymous> (:3000/static/js/main.chunk.js:1552)
at Module../src/components/render.js (:3000/static/js/main.chunk.js:2597)
at __webpack_require__ (:3000/static/js/bundle.js:785)
at fn (:3000/static/js/bundle.js:151)
at Module../src/components/FaceFilter.js (:3000/static/js/main.chunk.js:148)
at __webpack_require__ (:3000/static/js/bundle.js:785)
at fn (:3000/static/js/bundle.js:151)
at Module../src/components/App.js (:3000/static/js/main.chunk.js:72)
at __webpack_require__ (:3000/static/js/bundle.js:785)
at fn (:3000/static/js/bundle.js:151)
at Module../src/index.js (:3000/static/js/main.chunk.js:2627)
at __webpack_require__ (:3000/static/js/bundle.js:785)
at fn (:3000/static/js/bundle.js:151)
at Object.1 (:3000/static/js/main.chunk.js:2873)
at __webpack_require__ (:3000/static/js/bundle.js:785)
at checkDeferredModules (:3000/static/js/bundle.js:46)
at Array.webpackJsonpCallback [as push] (:3000/static/js/bundle.js:33)
at :3000/static/js/main.chunk.js:1
从201601到202008年我有60个数据点。
这是我的模式：
prophet model.我只用60分就对相同的数据进行训练和测试。
现在要提一个重要的问题，我想在holidays = pd.DataFrame({
"holiday":"New_Year","ds": pd.to_datetime(["2015-11-30","2016-11-30","2017-11-30","2018-11-30","2019-11-30","2020-11-30"]),"lower_window":-15,"upper_window":90
})
m = Prophet(interval_width=0.95,holidays=holidays)
for c in cols:
m.add_regressor(c)
m.fit(data)
train_forecast = m.predict(data)
当前方法：我采用yhat的均值和标准差，然后在这60个数据点上绘制1std，2std和3std线。无论超出第三标准，我都将其标记为预警。这是正确的方法吗？
问题1 ：由于我采用了interval_width = 0.95，所以我对未来数据的预测是否始终落在3个标准差的预测值之内？在那种情况下，我将永远不会异常。
我也在许多博客中看到了这种方法：
202009 to 202012但是对于看不见的数据（未来6个月），我不会使用y_original > forecasted['yhat_upper'] = 1
y_original< forecasted['yhat_lower']= -1
吗？因此，该方法不适用于预测值。
问题2：：如何使用现有的Prophet模型为未来的预测值标记异常？
解决方法
暂无找到可以解决该程序问题的有效方法，小编努力寻找整理中！
如果你已经找到好的解决方法，欢迎将解决方案带上本链接一起发送给小编。
小编邮箱:dio#foxmail.com (将#修改为@）