mirror of
https://github.com/ION606/COGMOD-HWI.git
synced 2026-05-14 22:16:57 +00:00
71 lines
2.3 KiB
Plaintext
71 lines
2.3 KiB
Plaintext
|
To solve Problem 5, follow these steps:
|
|||
|
|
|
|||
|
|
### Step 1: Simulate Data
|
|||
|
|
|
|||
|
|
### Step 2: Stan Model Code
|
|||
|
|
Write the Stan model (`bayesian_regression.stan`):
|
|||
|
|
```stan
|
|||
|
|
data {
|
|||
|
|
int<lower=0> N;
|
|||
|
|
vector[N] x;
|
|||
|
|
vector[N] y;
|
|||
|
|
}
|
|||
|
|
parameters {
|
|||
|
|
real alpha;
|
|||
|
|
real beta;
|
|||
|
|
real<lower=0> sigma_sq;
|
|||
|
|
}
|
|||
|
|
transformed parameters {
|
|||
|
|
real<lower=0> sigma = sqrt(sigma_sq);
|
|||
|
|
}
|
|||
|
|
model {
|
|||
|
|
sigma_sq ~ inv_gamma(1, 1); // Prior on variance
|
|||
|
|
alpha ~ normal(0, 10);
|
|||
|
|
beta ~ normal(0, 10);
|
|||
|
|
y ~ normal(alpha + beta * x, sigma); // Likelihood
|
|||
|
|
}
|
|||
|
|
```
|
|||
|
|
|
|||
|
|
### Step 3: Fit the Model and Check Diagnostics
|
|||
|
|
Use `pystan` or `cmdstanpy` to run the model. Check Rhat (≈1) and ESS (sufficiently large). For example:
|
|||
|
|
```python
|
|||
|
|
import cmdstanpy
|
|||
|
|
|
|||
|
|
model = cmdstanpy.CmdStanModel(stan_file="bayesian_regression.stan")
|
|||
|
|
data = {"N": N, "x": x, "y": y}
|
|||
|
|
fit = model.sample(data=data, chains=4, iter_sampling=2000)
|
|||
|
|
|
|||
|
|
# Check diagnostics
|
|||
|
|
print(fit.diagnose())
|
|||
|
|
```
|
|||
|
|
|
|||
|
|
### Step 4: Analyze Results for N=100
|
|||
|
|
Posterior summaries:
|
|||
|
|
- **Posterior means** should be close to true values (α=2.3, β=4.0, σ=2.0).
|
|||
|
|
- **Uncertainty**: Compute 95% credible intervals. Example output:
|
|||
|
|
- α: 2.1 ± 0.4 (1.7 to 2.5)
|
|||
|
|
- β: 3.8 ± 0.5 (3.3 to 4.3)
|
|||
|
|
- σ: 1.9 ± 0.2 (1.7 to 2.1)
|
|||
|
|
|
|||
|
|
### Step 5: Repeat with N=1000
|
|||
|
|
Increase sample size and rerun:
|
|||
|
|
```python
|
|||
|
|
N_large = 1000
|
|||
|
|
x_large = np.random.normal(size=N_large)
|
|||
|
|
y_large = alpha_true + beta_true * x_large + sigma_true * np.random.normal(size=N_large)
|
|||
|
|
```
|
|||
|
|
Fit the model again. Results will show:
|
|||
|
|
- **Tighter credible intervals** (e.g., β: 3.95 ± 0.1).
|
|||
|
|
- Reduced posterior variance, indicating higher precision.
|
|||
|
|
|
|||
|
|
### Key Observations:
|
|||
|
|
1. **Accuracy**: Posterior means align closely with true parameters.
|
|||
|
|
2. **Uncertainty**: Credible intervals narrow as \(N\) increases, reflecting reduced uncertainty.
|
|||
|
|
3. **Diagnostics**: Ensure Rhat ≈1 and sufficient ESS for reliable inferences.
|
|||
|
|
|
|||
|
|
**Visualization**: Plot prior vs. posterior histograms for parameters (using tools like `arviz` or `seaborn`), showing posterior concentration around true values, especially for \(N=1000\).
|
|||
|
|
|
|||
|
|
---
|
|||
|
|
|
|||
|
|
**Answer for LMS Submission**
|
|||
|
|
Implement the steps above, ensuring your write-up includes code snippets, diagnostic results, and graphical comparisons. Highlight the reduction in posterior variance when increasing \(N\), demonstrating the influence of data quantity on Bayesian inference.
|