Hi, vectorized data can be used with CNN directly. But, it is necessary to know whether the vectorization is correct or not. It should be adapted to the input dimensions required by CNN and further processing for the required output dimension as well. The case that you mention is close to the example shown in the following link (CNN applied to vectroized inputs for time series data):

https://machinelearningmastery.com/how-to-develop-convolutional-neural-network-models-for-time-series-forecasting/

If you need some articles as well, try the following:

1) Article Multi-Scale Convolutional Neural Networks for Time Series Cl...

2) https://www.aaai.org/ocs/index.php/IJCAI/IJCAI15/paper/view/10710/11297

If you are using MATLAB instead of Python, the core concept remains the same. It is mainly the syntax that differs slightly. Good luck!