# Prediction concept in Pymc3

**URL:** <https://discourse.pymc.io/t/prediction-concept-in-pymc3/7990>\
**Category:** Questions\
**Created:** [September 7, 2021, 9:23pm UTC](https://discourse.pymc.io/t/prediction-concept-in-pymc3/7990 "2021-09-07T21:23:18Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![GMCobraz\_T](https://yyz2.discourse-cdn.com/flex036/user_avatar/discourse.pymc.io/gmcobraz_t/32/4165_2.png) [@GMCobraz\_T](https://discourse.pymc.io/u/GMCobraz_T)\
**Post date:** [September 7, 2021, 9:23pm UTC](https://discourse.pymc.io/t/prediction-concept-in-pymc3/7990/1 "2021-09-07T21:23:19Z")

</div>

Good day.

I also read about out of sample prediction post as per link below:  
[https://docs.pymc.io/notebooks/posterior\_predictive.html](https://docs.pymc.io/notebooks/posterior_predictive.html)

I saw the example for out of sample prediction is changing the value of x and checking the output of y.

For example in the stock price prediction case, x is time while y is changes in price.

Based on my understanding, MCMC is ‘memoryless’ where future value is based on current value.

I am interested in instead of using time as x value, I want to know how to use current y value as input and check what is the predicted value.

May I know is there an example or some method to do it in Pymc3? using current y value in set\_data and run the sample posterior predictive to get the predicted y\_value

Thank you very much.