# Can I come back to analyzing posterior of a model after csv's have been dropped?

**URL:** <https://discourse.mc-stan.org/t/can-i-come-back-to-analyzing-posterior-of-a-model-after-csvs-have-been-dropped/25019>\
**Category:** General\
**Tags:** cmdstanpy\
**Created:** [October 28, 2021, 10:05am UTC](https://discourse.mc-stan.org/t/can-i-come-back-to-analyzing-posterior-of-a-model-after-csvs-have-been-dropped/25019 "2021-10-28T10:05:19Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![Jordan\_Howell](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.mc-stan.org/jordan_howell/32/4786_2.png) [@Jordan\_Howell](https://discourse.mc-stan.org/u/Jordan_Howell)\
**Post date:** [October 28, 2021, 10:05am UTC](https://discourse.mc-stan.org/t/can-i-come-back-to-analyzing-posterior-of-a-model-after-csvs-have-been-dropped/25019/1 "2021-10-28T10:05:19Z")

</div>

Hello.

I have fit a model and generated posterior predictive samples using cmdstanpy. Now that it’s done and the four csvs with samples are dropped to my computer, is there a standard practice on coming back to them later for analysis?

Is there a command to have cmdstanpy recognize them or pull them back into the environment once started back up?

---

<div class="post-metadata">

**Author:** ![yizhang](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.mc-stan.org/yizhang/32/15673_2.png) [@yizhang](https://discourse.mc-stan.org/u/yizhang)\
**Post date:** [November 10, 2021, 5:13pm UTC](https://discourse.mc-stan.org/t/can-i-come-back-to-analyzing-posterior-of-a-model-after-csvs-have-been-dropped/25019/2 "2021-11-10T17:13:11Z")

</div>

I’m a newbie to cmdstanpy too. It looks like [`from_csv`](https://mc-stan.org/cmdstanpy/api.html#from-csv) reads csv and creates a fit object.

---

<div class="post-metadata">

**Author:** ![WardBrian](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.mc-stan.org/wardbrian/32/12078_2.png) [@WardBrian](https://discourse.mc-stan.org/u/WardBrian)\
**Post date:** [November 10, 2021, 7:42pm UTC](https://discourse.mc-stan.org/t/can-i-come-back-to-analyzing-posterior-of-a-model-after-csvs-have-been-dropped/25019/3 "2021-11-10T19:42:01Z")

</div>

> [@yizhang](#):
>
> I’m a newbie to cmdstanpy too. It looks like [`from_csv`](https://mc-stan.org/cmdstanpy/api.html#from-csv) reads csv and creates a fit object.

This is correct!

Also, worth noting that by default cmdstanpy saves csv files to a temporary directory which is deleted later. You can save them in a more permanent place by either specifying an output directory during the `sample()` (or other method) call, or with `save_csvfiles()` after the fact

Here is the [relevant documentation](https://cmdstanpy.readthedocs.io/en/v1.0.0rc1/api.html#cmdstanpy.from_csv)