Rc 1.0 update by ihosten · Pull Request #9 · dierickxsimon/lactopy

ihosten · 2025-07-21T10:18:27Z

implementation of lt1 identification method based on standard increment
implementation of lt1 identification method based on standard increment starting from the lowest registered lactate measurement
start of implementation of the log-log method --> still in the works

…est outcomes

dierickxsimon · 2025-07-23T17:45:09Z

src/lactopy/lactate_models/lt1_lt2_breakpoint.py

+        pw_fit = piecewise_regression.Fit(self.X, self.y, n_breakpoints=2)
+        pw_results = pw_fit.get_results()
+
+        if "breakpoint1" not in pw_results.get("estimates", {}):


maybe also a good idea to add "breakpoint2"

done with elif

dierickxsimon · 2025-07-23T17:47:49Z

src/lactopy/lactate_models/lt1_lt2_breakpoint.py

+        self.y_raw_for_plot = y
+
+        # Ensure working on copies
+        self.X = np.asarray(copy.deepcopy(X))


no need. is already implemented in the base class. np.array returns a copy of the data and not a reference.

modified (also for loglog, lt1_si_lowest, lt1_si, lt2 breakpoint

dierickxsimon · 2025-07-23T17:49:06Z

src/lactopy/lactate_models/lt1_loglog.py

+
+
+class LT1_loglog(BaseModel):
+    """


this class does not contain logic for the log log method

modified the loglog and seems to work and give an acceptable outcome but peerreview is absolutely needed on this one!!!!!!

dierickxsimon · 2025-07-23T17:49:54Z

src/lactopy/lactate_models/lt1_lt2_breakpoint.py

+        self,
+        X: ArrayLike,
+        y: ArrayLike,
+        si: float = 0.5,


si is not used inside this function

same for other classes.

dierickxsimon · 2025-07-23T17:53:12Z

src/lactopy/lactate_models/lt1_lt2_breakpoint.py

+            si: the standard increment, defaults to 0.5
+
+            threshold_above_baseline (float, optional): Threshold above baseline
+            for filtering in the "modified" implementation. Defaults to 0.5.


no modified implementation inside this class

dierickxsimon · 2025-07-23T17:57:00Z

src/lactopy/lactate_models/lt1_si.py

+        if si > 0 :
+            X, y = copy.deepcopy(X), copy.deepcopy(y)
+            y_lt1 = y[0] + si
+            self.y_lt1 = y_lt1


good idea to add the lactate threshold as a instance based variable but should probably be defined in the base class.

it's just a naming to keep the function easy to read, it uses the predefined working of the base class

dierickxsimon · 2025-07-23T17:57:27Z

src/lactopy/lactate_models/lt1_si.py

+        else:
+            raise ValueError(f"Impossible lactate change to find LT1: {si}")
+
+        self.X = X


dierickxsimon · 2025-07-23T18:02:12Z

src/lactopy/lactate_models/lt2_breakpoint.py

+from lactopy.plots.lt2_breakpoint_plot import LT2_breakpoint_Plot
+
+
+class LT2_breakpoint(BaseModel):


is not needed already implemented in the LT1_LT2 breakpoint class.

true, was just a seperate extra can be deleted if we want, might be usefull for only determining 1 of the tresholds

dierickxsimon · 2025-07-23T18:03:19Z

src/lactopy/plots/lt1_loglog_plot.py

+        Plot the lt1_loglog model fit.
+        """
+        ax = super().plot_fit()
+        ax.set_title("LT1 Log-Log Model Fit")


I'm not sure if we need seprate titles for different plots. if so we should implement this in the base class and not in the all childeren to avoid tight coupling with maptlot lib

^might be interesting yes

dierickxsimon · 2025-07-23T18:05:18Z

src/lactopy/plots/lt1_loglog_plot.py

+        ax.set_title("LT1 Log-Log Model Fit")
+        return ax
+
+    def plot_predictions(self):


this is a rather reoccuring function in differen classes. i think we should add this logic in the base class

dierickxsimon · 2025-07-23T19:53:41Z

docs/quickstart.ipynb

the quickstart notebook is for the docs only change it for docs. if you want to create some test notebooks create one in notebooks/trash/

…sses patch for base classes

…g method using the piecewise regression module

…ed determination of lt1

…curve breakpoints using piecewise regression

…2_breakpoint method but only returning and plotting LT

…the base class bot for fitting and plotting should probably be discussed together

dierickxsimon · 2025-07-31T12:10:46Z

just some general toughts we need to fix before we can merge this update:

make docstring accesible in de mkdocs
pytests
uncoupling with matplotlib is quick fixed (only tight coupling in base class) but should eventually be moved to a seprete graph builder. but we will moves this to a later update.
we should think about the return object of the prediction class to make it more clear wether the prediction is about the lt1 or lt2. in the lt1_lt2_breakpoint is now done with a named tuple but we should think about a more general solultion. (issue [FEAT] return object of prediction #12 )
personallyt im not that big of a fan of having lt1_lt2_breakpoint and a lt1_breakpoint class it think the lt1_breakpoint is not needed.
lt1_lt2_breakpoint and all other classes that use breakpoint should overwrite the base fit class. And change the plots so that the piecewise fitting is plotted and not the the general fitting methods.
the obla and the lt_1_si classes can also be merged to one
we should also think about custom exceptions for failed fittings that are special to the lactate models. so users can used these to use differen models. (can be done in an other PR) [FEAT] Custom exceptions #13

dierickxsimon and others added 9 commits May 26, 2025 21:58

bug fix for when taking the derivative two times in a row

d10bc2c

add dmax plot for base

d4890a6

add docs

15991b6

clean up imports

916075d

inital start of docs

bad4141

impl of mod dmax

bf4b20f

update docs

71a1612

addition of extra csv files containing example test data of varying t…

d96d5c0

…est outcomes

reset of .python version

d50de85

ihosten changed the base branch from main to rc-1.0 July 21, 2025 12:14

dierickxsimon self-requested a review July 21, 2025 12:44

dierickxsimon requested changes Jul 23, 2025

View reviewed changes

dierickxsimon marked this pull request as draft July 23, 2025 18:16

dierickxsimon requested changes Jul 23, 2025

View reviewed changes

clean up

b1d8dbe

ihosten requested a review from dierickxsimon July 30, 2025 18:01

dierickxsimon and others added 9 commits July 30, 2025 20:15

Merge pull request #11 from dierickxsimon/cleanup/cleanup_of_base_cla…

dedd53d

…sses patch for base classes

add nb pre-commit to resolve merging errors with nb

bc6d05b

addition of the LT1 standard increment methods and start of the loglo…

9425826

…g method using the piecewise regression module

Workable version of the Log-Log method

7adc8a0

2 breakpoint detection instead of 1 in piecwise regression for improv…

750242b

…ed determination of lt1

implemented a method to calculate both lt1 and lt2 together based on …

5180fcc

…curve breakpoints using piecewise regression

implemented a breakpoint method to determine LT2, identical to lt1_lt…

33dd5d4

…2_breakpoint method but only returning and plotting LT

Processed mosed of the FB in the peer review, options for redefining …

2c967d6

…the base class bot for fitting and plotting should probably be discussed together

add nb pre-commit to resolve merging errors with nb

ac946e7

dierickxsimon force-pushed the rc-1.0_update branch from e80a7e2 to ac946e7 Compare July 30, 2025 18:52

ihosten marked this pull request as ready for review July 30, 2025 18:55

update code to work with the base class patch

2f90685

dierickxsimon marked this pull request as draft July 31, 2025 12:11

Base automatically changed from rc-1.0 to main October 29, 2025 15:09

		from lactopy.plots.lt2_breakpoint_plot import LT2_breakpoint_Plot


		class LT2_breakpoint(BaseModel):

Comments

Conversation

ihosten commented Jul 21, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

dierickxsimon commented Jul 31, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

dierickxsimon commented Jul 31, 2025 •

edited

Loading