feat!: remove all overdue deprecated APIs (Optimization/Results, FlowSystem methods, kwarg shims)

.github/workflows/tests.yaml

@@ -74,8 +74,8 @@ jobs:
       - name: Run tests
         run: |
           if [[ "${{ github.event.pull_request.draft }}" == "true" ]]; then
-            # Draft PR: skip examples, slow, and deprecated_api
-            pytest -v --numprocesses=auto -m "not examples and not slow and not deprecated_api"
+            # Draft PR: skip examples and slow tests
+            pytest -v --numprocesses=auto -m "not examples and not slow"
           else
             # Ready PR & main push: examples excluded via addopts
             pytest -v --numprocesses=auto

.gitignore

@@ -14,3 +14,6 @@ temp-plot.html
 site/
 *.egg-info
 uv.lock
+
+# test artifacts written to cwd
+fs.json

CHANGELOG.md

@@ -6,11 +6,25 @@ For more details regarding the individual PRs and contributors, please refer to
 
 !!! tip
 
+    If upgrading from v7.x, see the [Migration Guide v8](https://flixopt.github.io/flixopt/latest/user-guide/migration-guide-v8/).
     If upgrading from v5.x, see the [Migration Guide v6](https://flixopt.github.io/flixopt/latest/user-guide/migration-guide-v6/).
     If upgrading from v2.x, see the [v3.0.0 release notes](https://github.com/flixOpt/flixOpt/releases/tag/v3.0.0) and [Migration Guide v3](https://flixopt.github.io/flixopt/latest/user-guide/migration-guide-v3/).
 
 ---
 
+## [Unreleased]
+
+**Summary**: Removes all v4-era APIs that were deprecated in v5/v6 and scheduled for removal in v7.0.0. Code that runs warning-free on v7.x runs unchanged on v8. See the [Migration Guide v8](https://flixopt.github.io/flixopt/latest/user-guide/migration-guide-v8/).
+
+### 💥 Removed
+
+- **`Optimization` / `SegmentedOptimization` / `Results`** — use `flow_system.optimize(solver)` and read results via `flow_system.solution` / `flow_system.stats`
+- **`FlowSystem.sel()` / `isel()` / `resample()` / `coords`** — use `flow_system.transform.sel/isel/resample()` and `flow_system.indexes`
+- **`FlowSystem.plot_network()` / `network_infos()` / `start_network_app()` / `stop_network_app()`** and the PyVis-based `topology.plot_legacy()` — use `flow_system.topology.plot()` / `infos()` / `start_app()` / `stop_app()`; **pyvis is no longer a dependency**
+- **`FlowSystem.from_old_results()`** — re-run old optimizations with the current API (`from_old_dataset()` still loads pre-v5 configuration files and no longer warns)
+- **`Bus(excess_penalty_per_flow_hour=...)`** — now raises `TypeError`; use `imbalance_penalty_per_flow_hour`
+- **`normalize_weights` parameter** on `create_model` / `build_model` / `optimize()` — weights are always normalized
+
 ## [7.2.1](https://github.com/flixOpt/flixopt/compare/v7.2.0...v7.2.1) (2026-07-21)
 
 

docs/user-guide/glossary.md

@@ -88,7 +88,7 @@ Key concepts and terminology used throughout flixOpt.
 | **uptime** | Continuous duration equipment operates. Can be constrained with `min_uptime`, `max_uptime` in StatusParameters. |
 | **downtime** | Continuous duration equipment is off. Can be constrained with `min_downtime`, `max_downtime` in StatusParameters. |
 | **flow_hours** | Total energy delivered by a flow: sum of flow_rate * timestep_duration. Can be constrained with `flow_hours_min`, `flow_hours_max`. |
-| **excess_penalty** | Penalty applied when bus has more supply than demand. Soft constraint alternative to strict balance. |
+| **imbalance_penalty** | Penalty applied when a bus balance is violated (virtual supply/demand). Soft constraint alternative to strict balance. |
 | **shortage_penalty** | Penalty applied when bus has more demand than supply (unmet demand). |
 
 ## Weights and Aggregation

docs/user-guide/mathematical-notation/effects-and-dimensions.md

@@ -77,7 +77,7 @@ One effect is the **objective** (minimized). Others are tracked or constrained.
 
     ```python
     fx.Bus(
-        excess_penalty_per_flow_hour=1e6,    # Penalty for excess
+        imbalance_penalty_per_flow_hour=1e6,    # Penalty for excess
         shortage_penalty_per_flow_hour=1e6,  # Penalty for shortage
     )
     ```
@@ -259,7 +259,7 @@ A built-in `Penalty` effect enables soft constraints and prevents infeasibility:
 
 ```python
 fx.StatusParameters(effects_per_startup={'Penalty': 1})
-fx.Bus(label='heat', excess_penalty_per_flow_hour=1e5)
+fx.Bus(label='heat', imbalance_penalty_per_flow_hour=1e5)
 ```
 
 Penalty is weighted identically to the objective effect across all dimensions.

docs/user-guide/migration-guide-v5.md

@@ -246,20 +246,22 @@ boiler_vars = [v for v in flow_system.solution.data_vars if 'Boiler' in v]
     solution = xr.open_dataset('results/solution.nc4')
     ```
 
-### Migrating Old Result Files
+### Migrating Old Files
 
-If you have result files saved with the old API (v4.x), you can migrate them to the new format using `FlowSystem.from_old_results()`. This method:
+If you have FlowSystem configurations saved with the old API (v4.x), load them with `FlowSystem.from_old_dataset()`. This method:
 
-- Loads the old multi-file format (`*--flow_system.nc4`, `*--solution.nc4`)
+- Loads the old `*--flow_system.nc4` file
 - Renames deprecated parameters in the FlowSystem structure (e.g., `on_off_parameters` → `status_parameters`)
-- Attaches the solution data to the FlowSystem
+- Returns a FlowSystem without solution data
+
+Old *result* files can no longer be loaded (`from_old_results()` was removed in v8) — re-run the optimization with the current API instead.
 
 ```python
-# Load old results
-flow_system = fx.FlowSystem.from_old_results('results_folder', 'my_model')
+# Load an old FlowSystem configuration
+flow_system = fx.FlowSystem.from_old_dataset('results_folder/my_model--flow_system.nc4')
 
-# Access basic solution data (flow rates, sizes, charge states, etc.)
-flow_system.solution['Boiler(Q_th)|flow_rate'].plot()
+# Re-optimize with the current API
+flow_system.optimize(solver)
 
 # Save in new single-file format
 flow_system.to_netcdf('results/my_model_migrated.nc4')
@@ -396,7 +398,7 @@ stats.total_effects['costs'].groupby('component_type').sum()
 | **Replace Optimization class** | Use `flow_system.optimize(solver)` instead |
 | **Update results access** | Use `flow_system.solution['var_name']` pattern |
 | **Update I/O code** | Use `to_netcdf()` / `from_netcdf()` |
-| **Migrate old result files** | Use `FlowSystem.from_old_results(folder, name)` |
+| **Migrate old files** | Use `FlowSystem.from_old_dataset(path)` for configurations; re-run optimizations for results |
 | **Update transform methods** | Use `flow_system.transform.sel/isel/resample()` instead |
 | **Test thoroughly** | Verify results match v4.x outputs |
 | **Remove deprecated imports** | Remove `fx.Optimization`, `fx.Results` |

docs/user-guide/migration-guide-v7.md

@@ -9,6 +9,11 @@
     The `transform.cluster()` call signature is unchanged — only the `Clustering`
     result object and a few helpers changed.
 
+!!! info "Upgrading to v8?"
+    v8.0.0 removes the long-deprecated v4-era APIs (`Optimization`/`Results`,
+    `FlowSystem.sel/isel/resample`, PyVis network plotting). See the
+    [Migration Guide v8](migration-guide-v8.md).
+
 ---
 
 ## Overview

docs/user-guide/migration-guide-v8.md

@@ -0,0 +1,135 @@
+# Migration Guide: v7.x → v8.0.0
+
+!!! tip "Quick Start"
+    ```bash
+    pip install --upgrade flixopt
+    ```
+    v8.0.0 removes the long-deprecated v4-era APIs. If your code runs on v7.x
+    **without `DeprecationWarning`s**, it will run unchanged on v8.0.0 — this
+    guide is only relevant if you still use the old entry points below.
+
+---
+
+## Overview
+
+Every API removed in v8.0.0 has been deprecated since v5/v6 and has emitted a
+`DeprecationWarning` pointing to its replacement ever since. The replacements
+are unchanged — nothing new to learn, only old spellings to drop:
+
+| Removed | Replacement |
+|---------|-------------|
+| `fx.Optimization`, `fx.SegmentedOptimization` | `flow_system.optimize(solver)` or `build_model()` + `solve(solver)` |
+| `fx.results` / `Results` classes | `flow_system.solution`, `flow_system.stats` |
+| `FlowSystem.sel()` / `isel()` / `resample()` | `flow_system.transform.sel()` / `isel()` / `resample()` |
+| `FlowSystem.coords` | `flow_system.indexes` |
+| `FlowSystem.plot_network()`, `network_infos()` | `flow_system.topology.plot()`, `topology.infos()` |
+| `start_network_app()` / `stop_network_app()` | `topology.start_app()` / `topology.stop_app()` |
+| `topology.plot_legacy()` (PyVis) | `topology.plot()` (Plotly) |
+| `FlowSystem.from_old_results()` | re-run the optimization with the current API |
+| `Bus(excess_penalty_per_flow_hour=...)` | `Bus(imbalance_penalty_per_flow_hour=...)` |
+| `optimize(..., normalize_weights=...)` | remove the argument — weights are always normalized |
+
+---
+
+## 💥 Breaking Changes in v8.0.0
+
+### Optimization & Results classes removed
+
+The pre-v5 workflow objects are gone. Solve directly on the `FlowSystem` and
+read results from it:
+
+=== "v7.x and earlier (removed)"
+    ```python
+    import flixopt as fx
+
+    optimization = fx.Optimization('my_run', flow_system)
+    optimization.do_modeling()
+    optimization.solve(fx.solvers.HighsSolver())
+
+    results = fx.results.Results.from_optimization(optimization)
+    results.flow_rates()
+    ```
+
+=== "v8.0.0"
+    ```python
+    import flixopt as fx
+
+    flow_system.optimize(fx.solvers.HighsSolver())
+
+    flow_system.solution['Boiler(Q_th)|flow_rate']
+    flow_system.stats.flow_rates
+    ```
+
+`SegmentedOptimization` is removed without a direct replacement yet; a new
+segmented (rolling-horizon) API is planned.
+
+### Data methods moved to the `transform` accessor
+
+=== "v7.x and earlier (removed)"
+    ```python
+    fs_jan = flow_system.sel(time='2020-01')
+    fs_2h = flow_system.resample('2h', method='mean')
+    ```
+
+=== "v8.0.0"
+    ```python
+    fs_jan = flow_system.transform.sel(time='2020-01')
+    fs_2h = flow_system.transform.resample('2h', method='mean')
+    ```
+
+### Network visualization is Plotly-only
+
+`plot_network()`, the PyVis-based `topology.plot_legacy()`, and the network
+app wrappers on `FlowSystem` are removed. **PyVis is no longer a dependency.**
+
+=== "v7.x and earlier (removed)"
+    ```python
+    flow_system.plot_network(show=True)
+    flow_system.start_network_app()
+    ```
+
+=== "v8.0.0"
+    ```python
+    flow_system.topology.plot()
+    flow_system.topology.start_app()
+    ```
+
+### Renamed `Bus` argument is no longer bridged
+
+Passing the old name now raises a `TypeError` instead of warning:
+
+=== "v7.x and earlier (removed)"
+    ```python
+    fx.Bus('Heat', excess_penalty_per_flow_hour=1e5)
+    ```
+
+=== "v8.0.0"
+    ```python
+    fx.Bus('Heat', imbalance_penalty_per_flow_hour=1e5)
+    ```
+
+### Old result files can no longer be loaded
+
+`FlowSystem.from_old_results()` (the loader for pre-v5 `*--flow_system.nc4` +
+`*--solution.nc4` result pairs) is removed — re-run those optimizations with
+the current API.
+
+!!! info "Old *configuration* files still load"
+    `FlowSystem.from_old_dataset()` remains fully supported (and no longer warns): it loads a pre-v5
+    `*--flow_system.nc4` configuration, renames old parameters, and returns a
+    `FlowSystem` ready to re-optimize. Files saved with v5+
+    (`to_netcdf`/`from_netcdf`) are unaffected by any of this.
+
+    `from_old_dataset()` is planned for removal in **v9** — migrate pre-v5
+    files to the current single-file format before then.
+
+---
+
+## Checklist
+
+1. Run your project on v7.x with warnings visible:
+   `python -W once::DeprecationWarning your_script.py`
+2. Fix every `DeprecationWarning` using the table above.
+3. Upgrade: `pip install --upgrade flixopt`.
+
+If step 1 is silent, v8.0.0 is a drop-in upgrade.

docs/user-guide/optimization/index.md

@@ -363,12 +363,12 @@ If your model has no feasible solution:
 1. **Enable excess penalties on buses** to allow balance violations:
    ```python
    # Allow imbalance with high penalty cost (default is 1e5)
-   heat_bus = fx.Bus('Heat', excess_penalty_per_flow_hour=1e5)
+   heat_bus = fx.Bus('Heat', imbalance_penalty_per_flow_hour=1e5)
 
    # Or disable penalty to enforce strict balance
-   electricity_bus = fx.Bus('Electricity', excess_penalty_per_flow_hour=None)
+   electricity_bus = fx.Bus('Electricity', imbalance_penalty_per_flow_hour=None)
    ```
-   When `excess_penalty_per_flow_hour` is set, the optimization can violate bus balance constraints by paying a penalty, helping identify which constraints cause infeasibility.
+   When `imbalance_penalty_per_flow_hour` is set, the optimization can violate bus balance constraints by paying a penalty, helping identify which constraints cause infeasibility.
 
 2. **Use Gurobi for infeasibility analysis** - When using GurobiSolver and the model is infeasible, flixOpt automatically extracts and logs the Irreducible Inconsistent Subsystem (IIS):
    ```python

flixopt/__init__.py

@@ -17,7 +17,7 @@
 # - xr.Dataset.fxstats (from stats_accessor)
 import xarray_plotly as _xpx  # noqa: F401
 
-from . import clustering, linear_converters, plotting, results, solvers, tutorials
+from . import clustering, linear_converters, plotting, solvers, tutorials
 from . import stats_accessor as _fxstats  # noqa: F401
 from .carrier import Carrier, CarrierContainer
 from .comparison import Comparison
@@ -35,7 +35,6 @@
 from .elements import Bus, Flow
 from .flow_system import FlowSystem
 from .interface import InvestParameters, Piece, Piecewise, PiecewiseConversion, PiecewiseEffects, StatusParameters
-from .optimization import Optimization, SegmentedOptimization
 from .plot_result import PlotResult
 
 __all__ = [
@@ -55,8 +54,6 @@
     'LinearConverter',
     'Transmission',
     'FlowSystem',
-    'Optimization',
-    'SegmentedOptimization',
     'InvestParameters',
     'StatusParameters',
     'Piece',
@@ -66,7 +63,6 @@
     'PlotResult',
     'clustering',
     'plotting',
-    'results',
     'linear_converters',
     'solvers',
     'tutorials',

flixopt/elements.py

@@ -313,9 +313,6 @@ def __init__(
         **kwargs,
     ):
         super().__init__(label, meta_data=meta_data)
-        imbalance_penalty_per_flow_hour = self._handle_deprecated_kwarg(
-            kwargs, 'excess_penalty_per_flow_hour', 'imbalance_penalty_per_flow_hour', imbalance_penalty_per_flow_hour
-        )
         self._validate_kwargs(kwargs)
         self.carrier = carrier.lower() if carrier else None  # Store as lowercase string
         self.imbalance_penalty_per_flow_hour = imbalance_penalty_per_flow_hour