updated data_container.ipynb to reflect changes to pm.Data containers in pymc v5.16 #676

Dekermanjian · 2024-07-28T11:52:16Z

related to #659

The data container notebook is out of date. It uses the deprecated pm.ConstantData and pm.MutableData methods.
The changes I made were:

Update the introduction to reflect the change to using pm.Data but keeping a section that refers to the deprecated pm.MutableData and pm.ConstantData in case a new user sees that in example code elsewhere.
Replace all instances of pm.ConstantData and pm.MutableData with pm.Data in the code and in the markdown text
Replace the instance of mutable coords with coords to reflect that all coordinates are now mutable and explain that fact in the preceding markdown text
I removed some text that no longer applied like checking the type of data container twice previously for each type of data container
I made a couple of typo corrections

📚 Documentation preview 📚: https://pymc-examples--676.org.readthedocs.build/en/676/

review-notebook-app · 2024-07-28T11:52:21Z

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

twiecki · 2024-07-30T08:49:03Z

examples/fundamentals/data_container.myst.md

@@ -58,19 +58,17 @@ This notebook will illustrate each of these benefits in turn, and show you the b

 +++

-## Types of Data Containers
+## Current and past states of Data Containers


I think this whole section can be deleted and maybe replaced with an info box of your last comment.

I agree, I will make the change.

twiecki · 2024-07-30T08:49:58Z

examples/fundamentals/data_container.myst.md

@@ -99,20 +97,20 @@ Furthermore, inside `idata`, PyMC has automatically saved the observed (endogeno
 idata.observed_data
 ```

-In this next model, we create a `pm.ConstantData` containers to hold the observations, and pass this container to the `observed`. We also make a `pm.ConstantData` container to hold the `x` data:
+In this next model, we create a `pm.Data` container to hold the observations, and pass this container to the `observed`. We also make a `pm.Data` container to hold the `x` data:


Suggested change

In this next model, we create a `pm.Data` container to hold the observations, and pass this container to the `observed`. We also make a `pm.Data` container to hold the `x` data:

In this next model, we create a {class}`pm.Data` container to hold the observations, and pass this container to the `observed`. We also make a {class}`pm.Data` container to hold the `x` data:

Change everywhere we use pm.Data.

Thank you for clarifying that for me. I was not sure if the {class} was only supposed to appear the first time we introduce the class. I will make the changes to fix it.

It should be one of {class}`pymc.Data` (will render pymc.Data as link) or {class}`~pymc.Data` (will render only Data as link). Sphinx doesn't know about pm. I prefer the 1st one.

@twiecki @OriolAbril Would it be preferable to change all pm.Data to {class}pymc.Data or is it preferred that I go back to keeping it as pm.Data without {class}? At the start of the notebook, there is a reference to {class}pymc.Data.

I personally prefer pymc.Data as a link, but I don't really care much either way

Okay, I went ahead and changed it so that it is {class} pymc.Data everywhere in the markdown text.

twiecki · 2024-07-30T08:50:16Z

examples/fundamentals/data_container.myst.md

    beta = pm.Normal("beta")
    mu = pm.Deterministic("mu", beta * x_data)
    sigma = pm.Exponential("sigma", 1)
    obs = pm.Normal("obs", mu=mu, sigma=sigma, observed=y_data)
    idata = pm.sample(random_seed=RANDOM_SEED)
 ```

-Because we used a `pm.ConstantData` container, the data now appears on our probabilistic graph. It is downstream from `obs` (since the `obs` variable "causes" the data), shaded in gray (because it is observed), and has a special rounded square shape to emphasize that it is data. We also see that `x_data` has been added to the graph.
+Because we used a `pm.Data` container, the data now appears on our probabilistic graph. It is downstream from `obs` (since the `obs` variable "causes" the data), shaded in gray (because it is observed), and has a special rounded square shape to emphasize that it is data. We also see that `x_data` has been added to the graph.


Suggested change

Because we used a `pm.Data` container, the data now appears on our probabilistic graph. It is downstream from `obs` (since the `obs` variable "causes" the data), shaded in gray (because it is observed), and has a special rounded square shape to emphasize that it is data. We also see that `x_data` has been added to the graph.

Because we used a `pm.Data` container, the data now appears in our probabilistic graph. It is downstream from `obs` (since the `obs` variable "causes" the data), shaded in gray (because it is observed), and has a special rounded square shape to emphasize that it is data. We also see that `x_data` has been added to the graph.

Yup, I missed that. Thank you, for catching it. I will fix it.

… to , fixed typo

twiecki · 2024-08-05T10:44:27Z

examples/fundamentals/data_container.myst.md

-
- In past versions of PyMC, the only data container was `pm.Data`. This container is still available for backwards compatability, but the current best practice is to use either `pm.MutableData` or `pm.ConstantData`. 
+:::{important}
+In past versions of PyMC, there were two types of data containers {func}`pymc.MutableData` and {func}`pymc.ConstantData` these are now deprecated.


Suggested change

In past versions of PyMC, there were two types of data containers {func}`pymc.MutableData` and {func}`pymc.ConstantData` these are now deprecated.

In past versions of PyMC, there were two types of data containers {func}`pymc.MutableData` and {func}`pymc.ConstantData`. These have been deprecated as all data containers are mutable now.

Thank you, for the review @twiecki. I have made the change that you have suggested.

twiecki · 2024-08-05T13:09:04Z

Thanks @Dekermanjian!

updated data_container.ipynb to reflect changes to pm.Data containers

dca848c

twiecki reviewed Jul 30, 2024

View reviewed changes

Dekermanjian added 2 commits July 30, 2024 04:55

replaced section with info box about deprecated data container, added…

2df0716

… to , fixed typo

swapped out class pm.Data to just pm.Data

bc7ab3c

OriolAbril linked an issue Aug 1, 2024 that may be closed by this pull request

Update data container notebook to reflect changes to pm.Data #659

Closed

OriolAbril mentioned this pull request Aug 1, 2024

fix typo #609

Closed

3 tasks

swapped pm.Data to class pymc.data in markdown text

c96081f

twiecki reviewed Aug 5, 2024

View reviewed changes

modified text around deprecated Mutable and Constant data containers

aea5349

twiecki merged commit 0a2ea64 into pymc-devs:main Aug 5, 2024
2 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

updated data_container.ipynb to reflect changes to pm.Data containers in pymc v5.16 #676

updated data_container.ipynb to reflect changes to pm.Data containers in pymc v5.16 #676

Dekermanjian commented Jul 28, 2024 •

edited by github-actions bot

Loading

review-notebook-app bot commented Jul 28, 2024

twiecki Jul 30, 2024

Dekermanjian Jul 30, 2024

twiecki Jul 30, 2024

twiecki Jul 30, 2024

Dekermanjian Jul 30, 2024

OriolAbril Jul 30, 2024

Dekermanjian Jul 30, 2024

OriolAbril Aug 1, 2024

Dekermanjian Aug 1, 2024

twiecki Jul 30, 2024

Dekermanjian Jul 30, 2024

twiecki Aug 5, 2024

Dekermanjian Aug 5, 2024

twiecki commented Aug 5, 2024

	In this next model, we create a `pm.Data` container to hold the observations, and pass this container to the `observed`. We also make a `pm.Data` container to hold the `x` data:
	In this next model, we create a {class}`pm.Data` container to hold the observations, and pass this container to the `observed`. We also make a {class}`pm.Data` container to hold the `x` data:

	Because we used a `pm.Data` container, the data now appears on our probabilistic graph. It is downstream from `obs` (since the `obs` variable "causes" the data), shaded in gray (because it is observed), and has a special rounded square shape to emphasize that it is data. We also see that `x_data` has been added to the graph.
	Because we used a `pm.Data` container, the data now appears in our probabilistic graph. It is downstream from `obs` (since the `obs` variable "causes" the data), shaded in gray (because it is observed), and has a special rounded square shape to emphasize that it is data. We also see that `x_data` has been added to the graph.

	In past versions of PyMC, there were two types of data containers {func}`pymc.MutableData` and {func}`pymc.ConstantData` these are now deprecated.
	In past versions of PyMC, there were two types of data containers {func}`pymc.MutableData` and {func}`pymc.ConstantData`. These have been deprecated as all data containers are mutable now.

updated data_container.ipynb to reflect changes to pm.Data containers in pymc v5.16 #676

updated data_container.ipynb to reflect changes to pm.Data containers in pymc v5.16 #676

Conversation

Dekermanjian commented Jul 28, 2024 • edited by github-actions bot Loading

review-notebook-app bot commented Jul 28, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

twiecki commented Aug 5, 2024

Dekermanjian commented Jul 28, 2024 •

edited by github-actions bot

Loading