I will forecast model time series using python

Complete analysis and prediction of Time series in my python pipeline:

The offer includes:

• Google Colab Notebook (web-app, doesn't require to install any software) containing the full python pipeline that reads data, generate the predictions, export the output
• Documentation including theory behind each graph & decision taken provided directly in the notebook.
• Replicability/Scalability: Analysis completely replicable by just changing input data
• Methodologies: Traditional (ARIMA) and/or new (FBProphet) methodologies including Machine Learning

The analysis consists of: 

1. Time Series model fitting: Fit a TS model to apply to past/future data
2. TS Decomposition into trend, Seasonal (yearly/weekly/daily/holidays) & Error components
3. Anomaly Detection: Identify anomalies in the past and corresponding importance
4. TS Predictions: Predictions for future data, along with corresponding Interval of Confidence
5. Diagnostics of predictions: Accuracy measures and errors of predictions (MAE, MSE, MAPE, etc)
6. (Only for ARIMA): ACF/PACF/ Moving Average decomposition, check stationarity
7. Multivariate modeling: Predict y using multivariate models that use other covariates (x1, x2, x3, etc.)