Add minimal pymc example #7281

HarshvirSandhu · 2024-04-26T20:45:52Z

Description

Example is taken from this notebook and made to look like the readme example of sunode
Not sure if the example is placed correctly.

Related Issue

Closes Include minimal pymc example in README #7267

Checklist

Checked that the pre-commit linting/style checks pass
Added necessary documentation (docstrings and/or example notebooks)

Type of change

Documentation

📚 Documentation preview 📚: https://pymc--7281.org.readthedocs.build/en/7281/

welcome · 2024-04-26T20:45:55Z

]
💖 Thanks for opening this pull request! 💖 The PyMC community really appreciates your time and effort to contribute to the project. Please make sure you have read our Contributing Guidelines and filled in our pull request template to the best of your ability.

twiecki · 2024-04-29T15:48:58Z

README.rst

+      mu = x @ betas
+
+      # Likelihood
+      y = pm.Normal("y", mu, sigma, dims=["trial"])


@ricardoV94 what might the output be? I think at least having a short story like "imagine we ran an experiment where for different levels of hardness, conducitivty and temperature, we measured Y".

Sounds good

No, I'm asking what we're measuring in this hypothetical experiment. Just to round out the example.

Pick one from ChatGPT:

Material Strength: This could include tensile strength, yield strength, or fatigue strength. These measures could be affected by changes in material hardness and temperature.

Electrical Properties: Besides conductivity, which is already a variable, YY might represent electrical resistance or capacitance, which can change with temperature and material properties.

Thermal Properties: Such as thermal expansion or specific heat capacity, which could be influenced by the material's hardness and its conductivity.

Optical Properties: Such as reflectivity or light absorption, which might change with temperature and the physical state of the material.

Chemical Reactivity: Rate of a chemical reaction that could be influenced by the material's properties and temperature.

Phase Changes: The point at which a material changes from solid to liquid (melting point) or from liquid to gas (boiling point), which can be influenced by the material's composition and environmental conditions.

Durability or Wear Resistance: How well a material can withstand wear or degradation over time, potentially influenced by its hardness and operating temperature.

Here's a different example from GPT that I think works better:

Objective: Investigate the effects of sunlight exposure, water amount, and soil nitrogen content on plant growth.

Background: Plant growth can be influenced by multiple factors, and understanding these relationships is crucial for optimizing agricultural practices. In this experiment, we aim to predict the growth of a plant based on three different environmental variables.

Experiment Setup:
Variables:
Independent Variables:
Sunlight Hours (X1): Number of hours the plant is exposed to sunlight daily.
Water Amount (X2): Daily water amount given to the plant (in milliliters).
Soil Nitrogen Content (X3): Percentage of nitrogen content in the soil.
Dependent Variable:
Plant Growth (Y): Measured as the increase in plant height (in centimeters) over a certain period.

Sounds good, just say data is standardized so the normal draws and likelihood make sense

@HarshvirSandhu Can you make these changes to the example?

twiecki · 2024-05-02T20:20:37Z

README.rst


   # Define coordinate values for all dimensions of the data
   coords={
    "trial": range(100),
-    "features": ["hardness", "conductivity", "temperature"],
+    "features": ["X1", "X2", "X3"],


Suggested change

"features": ["X1", "X2", "X3"],

"features": ["sunlight hours", "water amount", "soil nitrogen"],

twiecki · 2024-05-02T20:20:56Z

README.rst

@@ -45,11 +52,20 @@ Linear Regression Example
   x_dist = pm.Normal.dist(shape=(100, 3))
   x_data = pm.draw(x_dist, random_seed=seed)

+   # Independent Variables:
+   # Sunlight Hours (X1): Number of hours the plant is exposed to sunlight daily.


Suggested change

# Sunlight Hours (X1): Number of hours the plant is exposed to sunlight daily.

# Sunlight Hours: Number of hours the plant is exposed to sunlight daily.

twiecki · 2024-05-02T20:21:47Z

README.rst

+
+Plant growth can be influenced by multiple factors, and understanding these relationships is crucial for optimizing agricultural practices.
+
+In this experiment, we aim to predict the growth of a plant based on three different environmental variables.


Suggested change

In this experiment, we aim to predict the growth of a plant based on three different environmental variables.

Imagine we conduct an experiment to predict the growth of a plant based on three different environmental variables.

twiecki · 2024-05-02T20:21:57Z

README.rst

@@ -36,6 +36,13 @@ Features

 Linear Regression Example
 ==========================
+
+**Background**


Suggested change

**Background**

twiecki · 2024-05-02T20:23:27Z

README.rst

+      mu = x @ betas
+
+      # Likelihood
+      y = pm.Normal("y", mu, sigma, dims=["trial"])


Suggested change

y = pm.Normal("y", mu, sigma, dims=["trial"])

plant_growth = pm.Normal("plant growth (z-scored)", mu, sigma, dims="trial")

twiecki · 2024-05-06T11:49:18Z

README.rst

@@ -137,7 +135,7 @@ sigma                 0.511  0.037     0.438      0.575        0.001      0
   with pm.do(
    inference_model,
    {inference_model["betas"]: inference_model["betas"] * [1, 1, 0]},
-   ) as heat_death_model:
+   ) as new_model:


Suggested change

) as new_model:

) as plant_growth_model:

twiecki · 2024-05-06T11:50:43Z

README.rst

@@ -80,7 +79,7 @@ In this experiment, we aim to predict the growth of a plant based on three diffe
      mu = x @ betas

      # Likelihood
-      y = pm.Normal("y", mu, sigma, dims=["trial"])
+      plant_growth = pm.Normal("plant growth (z-scored)", mu, sigma, dims="trial")


Suggested change

plant_growth = pm.Normal("plant growth (z-scored)", mu, sigma, dims="trial")

# Assuming we measure deviation of each plant from baseline

plant_growth = pm.Normal("plant growth", mu, sigma, dims="trial")

ricardoV94 · 2024-05-09T08:17:34Z

README.rst

+         random_seed=seed,
+      )
+
+   pm.stats.summary(idata.predictions, kind="stats")


Shouldn't we show all the summaries outputs? Why only the first?

I think for starters it's TMI and can scare people off. Convergence diagnostics is more advanced than what we want to demo here.

Oh, I thought you meant more columns, but you meant more rows?

I mean't every time the have pm.stats.summary, we should show the output. I already removed the extra convergence columns with kind="stats". Right now it's only showing for the first usage

twiecki · 2024-05-12T15:38:19Z

README.rst

+==========================  ======  =====  ========  =========
+Output                       mean    sd     hdi_3%    hdi_97%
+==========================  ======  =====  ========  =========
+plant growth (z-scored)[0]  14.21   0.509    13.232     15.144


this is still the old name ("z-scored").

twiecki · 2024-05-13T09:36:36Z

README.rst

+==========================  ======  =====  ========  =========
+Output                       mean    sd     hdi_3%    hdi_97%
+==========================  ======  =====  ========  =========
+plant growth (z-scored)[0]  14.153  0.509    13.181     15.096


Also needs updated name.

kirahowe · 2024-06-15T15:07:44Z

Having a look at this at the PyData London 2024 hackathon 😄

kirahowe · 2024-06-15T15:36:29Z

Couldn't commit to this branch, so there's a PR here on another fork: #7358. As of today, the example works on my machine and the outputs reflect the names used in the example.

twiecki · 2024-06-15T15:52:02Z

Closing in favor of #7358.

Add minimal pymc example

10a91a2

twiecki reviewed Apr 29, 2024

View reviewed changes

Add experiment background info

f479281

twiecki reviewed May 2, 2024

View reviewed changes

Update variable names

7f66672

twiecki reviewed May 6, 2024

View reviewed changes

Update variable names

9addc5c

twiecki approved these changes May 8, 2024

View reviewed changes

ricardoV94 reviewed May 9, 2024

View reviewed changes

Add summary outputs

b2e5c1e

twiecki reviewed May 12, 2024

View reviewed changes

twiecki reviewed May 13, 2024

View reviewed changes

twiecki closed this Jun 15, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add minimal pymc example #7281

Add minimal pymc example #7281

HarshvirSandhu commented Apr 26, 2024 •

edited by github-actions bot

Loading

welcome bot commented Apr 26, 2024

twiecki Apr 29, 2024

ricardoV94 Apr 29, 2024

twiecki May 1, 2024

ricardoV94 May 1, 2024

twiecki May 1, 2024

ricardoV94 May 1, 2024 •

edited

Loading

twiecki May 2, 2024

HarshvirSandhu May 2, 2024

twiecki May 2, 2024

twiecki May 2, 2024

twiecki May 2, 2024

twiecki May 2, 2024

twiecki May 2, 2024

twiecki May 6, 2024 •

edited

Loading

twiecki May 6, 2024 •

edited

Loading

ricardoV94 May 9, 2024

twiecki May 12, 2024

twiecki May 12, 2024

ricardoV94 May 13, 2024 •

edited

Loading

twiecki May 12, 2024

twiecki May 13, 2024

kirahowe commented Jun 15, 2024

kirahowe commented Jun 15, 2024 •

edited

Loading

twiecki commented Jun 15, 2024

	"features": ["X1", "X2", "X3"],
	"features": ["sunlight hours", "water amount", "soil nitrogen"],

	# Sunlight Hours (X1): Number of hours the plant is exposed to sunlight daily.
	# Sunlight Hours: Number of hours the plant is exposed to sunlight daily.


		Plant growth can be influenced by multiple factors, and understanding these relationships is crucial for optimizing agricultural practices.

		In this experiment, we aim to predict the growth of a plant based on three different environmental variables.

	In this experiment, we aim to predict the growth of a plant based on three different environmental variables.
	Imagine we conduct an experiment to predict the growth of a plant based on three different environmental variables.

	y = pm.Normal("y", mu, sigma, dims=["trial"])
	plant_growth = pm.Normal("plant growth (z-scored)", mu, sigma, dims="trial")

	plant_growth = pm.Normal("plant growth (z-scored)", mu, sigma, dims="trial")
	# Assuming we measure deviation of each plant from baseline
	plant_growth = pm.Normal("plant growth", mu, sigma, dims="trial")

Add minimal pymc example #7281

Add minimal pymc example #7281

Conversation

HarshvirSandhu commented Apr 26, 2024 • edited by github-actions bot Loading

Description

Related Issue

Checklist

Type of change

welcome bot commented Apr 26, 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

ricardoV94 May 1, 2024 • edited Loading

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 May 6, 2024 • edited Loading

Choose a reason for hiding this comment

twiecki May 6, 2024 • edited Loading

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

ricardoV94 May 13, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kirahowe commented Jun 15, 2024

kirahowe commented Jun 15, 2024 • edited Loading

twiecki commented Jun 15, 2024

HarshvirSandhu commented Apr 26, 2024 •

edited by github-actions bot

Loading

ricardoV94 May 1, 2024 •

edited

Loading

twiecki May 6, 2024 •

edited

Loading

twiecki May 6, 2024 •

edited

Loading

ricardoV94 May 13, 2024 •

edited

Loading

kirahowe commented Jun 15, 2024 •

edited

Loading