feat: Normalized rows (#205)

* feat: add NormalizedRow API and ResidualErrorModel

This commit adds two related features:

## NormalizedRow API (parser/)
- New  struct for format-agnostic data parsing
- Decouples column mapping from event creation logic
- Full ADDL/II expansion support (both positive and negative directions)
- Refactors pmetrics.rs to use NormalizedRow internally
- Enables external tools (like vial) to reuse parsing logic without
  reimplementing ADDL expansion

## ResidualErrorModel (data/)
- New  for parametric algorithms (SAEM, FOCE)
- Uses prediction-based sigma (vs observation-based in ErrorModel)
- Adds  and  functions
- Documentation clarifying ErrorModel vs ResidualErrorModel usage

Both features are independent but included together to avoid merge conflicts.

Co-authored-by: Julián D. Otálvaro <juliandavid347@gmail.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Author: 3 people

CHANGELOG.md

@@ -11,7 +11,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Other
 
-- *(Exa)* when installing Vial on MacOs, the environment varaibles are not completly shared to the sandbox in which Vial is running, this changes are meant to provide vial a better way to approach finding the rust binary ([#181](https://github.com/LAPKB/pharmsol/pull/181))
+- _(Exa)_ when installing Papir on MacOs, the environment varaibles are not completly shared to the sandbox in which Papir is running, this changes are meant to provide papir a better way to approach finding the rust binary ([#181](https://github.com/LAPKB/pharmsol/pull/181))
 - Update diffsol requirement from =0.7.0 to =0.8.0 ([#176](https://github.com/LAPKB/pharmsol/pull/176))
 - Update criterion requirement from 0.7.0 to 0.8.0 ([#177](https://github.com/LAPKB/pharmsol/pull/177))
 - Update libloading requirement from 0.8.6 to 0.9.0 ([#162](https://github.com/LAPKB/pharmsol/pull/162))
@@ -218,7 +218,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 - Add events to occasions
 - Expose functions for number of states and outeqs
-- Add support for multiple error models  ([#65](https://github.com/LAPKB/pharmsol/pull/65))
+- Add support for multiple error models ([#65](https://github.com/LAPKB/pharmsol/pull/65))
 
 ## [0.9.1](https://github.com/LAPKB/pharmsol/compare/v0.9.0...v0.9.1) - 2025-05-22
 

README.md

@@ -87,7 +87,9 @@ let ode = equation::ODE::new(
 
 Analytical solutions provide 20-33× speedups compared to equivalent ODE formulations. See [benchmarks](benches/) for details.
 
-## Documentation
+
+
+# Links
 
 - [API Documentation](https://lapkb.github.io/pharmsol/pharmsol/)
 - [Examples](examples/)

examples/one_compartment.rs

@@ -56,23 +56,23 @@ fn main() -> Result<(), pharmsol::PharmsolError> {
     );
 
     // Define the error models for the observations
-    let ems = ErrorModels::new().
+    let ems = AssayErrorModels::new().
     // For this example, we use a simple additive error model with 5% error
     add(
         0,
-        ErrorModel::additive(ErrorPoly::new(0.0, 0.05, 0.0, 0.0), 0.0),
+        AssayErrorModel::additive(ErrorPoly::new(0.0, 0.05, 0.0, 0.0), 0.0),
     )?;
 
     // Define the parameter values for the simulations
     let ke = 1.022; // Elimination rate constant
     let v = 194.0; // Volume of distribution
 
     // Compute likelihoods and predictions for both models
-    let analytical_likelihoods = an.estimate_likelihood(&subject, &vec![ke, v], &ems, false)?;
+    let analytical_likelihoods = an.estimate_log_likelihood(&subject, &vec![ke, v], &ems, false)?;
 
     let analytical_predictions = an.estimate_predictions(&subject, &vec![ke, v])?;
 
-    let ode_likelihoods = ode.estimate_likelihood(&subject, &vec![ke, v], &ems, false)?;
+    let ode_likelihoods = ode.estimate_log_likelihood(&subject, &vec![ke, v], &ems, false)?;
 
     let ode_predictions = ode.estimate_predictions(&subject, &vec![ke, v])?;
 
@@ -81,7 +81,7 @@ fn main() -> Result<(), pharmsol::PharmsolError> {
     println!("│           │   Analytical    │       ODE       │     Difference      │");
     println!("├───────────┼─────────────────┼─────────────────┼─────────────────────┤");
     println!(
-        "│ Likelihood│ {:>15.6} │ {:>15.6} │ {:>19.2e} │",
+        "│ Log-Likeli│ {:>15.6} │ {:>15.6} │ {:>19.2e} │",
         analytical_likelihoods,
         ode_likelihoods,
         analytical_likelihoods - ode_likelihoods

src/data/builder.rs

@@ -66,7 +66,7 @@ impl SubjectBuilder {
     ///
     /// * `time` - Time of the bolus dose
     /// * `amount` - Amount of drug administered
-    /// * `input` - The compartment number (zero-indexed) receiving the dose
+    /// * `input` - The compartment number receiving the dose
     pub fn bolus(self, time: f64, amount: f64, input: usize) -> Self {
         let bolus = Bolus::new(time, amount, input, self.current_occasion.index());
         let event = Event::Bolus(bolus);
@@ -79,7 +79,7 @@ impl SubjectBuilder {
     ///
     /// * `time` - Start time of the infusion
     /// * `amount` - Total amount of drug to be administered
-    /// * `input` - The compartment number (zero-indexed) receiving the dose
+    /// * `input` - The compartment number receiving the dose
     /// * `duration` - Duration of the infusion in time units
     pub fn infusion(self, time: f64, amount: f64, input: usize, duration: f64) -> Self {
         let infusion = Infusion::new(time, amount, input, duration, self.current_occasion.index());
@@ -93,8 +93,7 @@ impl SubjectBuilder {
     ///
     /// * `time` - Time of the observation
     /// * `value` - Observed value (e.g., drug concentration)
-    /// * `outeq` - Output equation number (zero-indexed) corresponding to this observation
-    /// * `errorpoly` - Error polynomial coefficients (c0, c1, c2, c3)
+    /// * `outeq` - Output equation number corresponding to this observation
     pub fn observation(self, time: f64, value: f64, outeq: usize) -> Self {
         let observation = Observation::new(
             time,