Dynamic logging

pyproject.toml

@@ -9,7 +9,7 @@ profile = "black"
 
 [project]
 name = "turftopic"
-version = "0.27.0"
+version = "0.27.1"
 description = "Topic modeling with contextual representations from sentence transformers."
 authors = [
    { name = "Márton Kardos <power.up1163@gmail.com>", email = "martonkardos@cas.au.dk" }

turftopic/models/_snmf.py

@@ -172,18 +172,18 @@ def fit_new_components(self, X: np.ndarray, n_new_components: int):
         ):
             G, F, error = _step(G, F)
             difference = prev_error - error
-            if (error < self.error_at_init) and (
-                (prev_error - error) / self.error_at_init
-            ) < self.tol:
-                if self.verbose:
-                    print(f"Converged after {i} iterations")
-                self.n_iter_ = i
-                break
-            prev_error = error
-            if self.verbose:
-                print(
-                    f"Iteration: {i}, Error: {error}, init_error: {self.error_at_init}, difference from previous: {difference}"
-                )
+            # if (error < self.error_at_init) and (
+            #     (prev_error - error) / self.error_at_init
+            # ) < self.tol:
+            #     if self.verbose:
+            #         print(f"Converged after {i} iterations")
+            #     self.n_iter_ = i
+            #     break
+            # prev_error = error
+            # if self.verbose:
+            #     print(
+            #         f"Iteration: {i}, Error: {error}, init_error: {self.error_at_init}, difference from previous: {difference}"
+            #     )
         self.components_ = np.array(F.T)
         self.n_iter_ = i
         self.n_components = old_n_components + n_new_components

turftopic/models/senstopic.py

@@ -4,7 +4,7 @@
 
 import numpy as np
 from rich.console import Console
-from sentence_transformers import SentenceTransformer
+from rich.progress import track
 from sklearn.base import copy
 from sklearn.exceptions import NotFittedError
 from sklearn.feature_extraction.text import CountVectorizer
@@ -455,26 +455,39 @@ def fit_transform_dynamic(
             document_topic_matrix = self.transform(
                 raw_documents, embeddings=embeddings
             )
-        time_labels, self.time_bin_edges = self.bin_timestamps(
-            timestamps, bins
-        )
-        n_comp, n_vocab = self.components_.shape
-        n_bins = len(self.time_bin_edges) - 1
-        self.axial_temporal_components_ = np.full(
-            (n_bins, n_comp, n_vocab),
-            np.nan,
-            dtype=self.components_.dtype,
-        )
-        self.temporal_importance_ = np.zeros((n_bins, n_comp))
-        for i_timebin in np.unique(time_labels):
+        console = Console()
+        with console.status("Fitting temporally conditioned topics") as status:
+            status.update("Labelling documents based on time bins.")
+            time_labels, self.time_bin_edges = self.bin_timestamps(
+                timestamps, bins
+            )
+            console.log("Documents binned.")
+            status.update("Initializing components.")
+            n_comp, n_vocab = self.components_.shape
+            n_bins = len(self.time_bin_edges) - 1
+            self.axial_temporal_components_ = np.full(
+                (n_bins, n_comp, n_vocab),
+                np.nan,
+                dtype=self.components_.dtype,
+            )
+            console.log("Components initialized.")
+            self.temporal_importance_ = np.zeros((n_bins, n_comp))
+        for i_timebin in track(
+            np.unique(time_labels),
+            description="Calculating temporal components for each time slice.",
+            console=console,
+        ):
             t_dt = document_topic_matrix[time_labels == i_timebin]
             t_X = self.embeddings[time_labels == i_timebin]
             t_imp, t_comp = self._fit_timebin(t_X, t_dt)
             self.temporal_importance_[i_timebin, :] = t_imp
             self.axial_temporal_components_[i_timebin, :, :] = t_comp
-        self.estimate_components(
-            self.feature_importance,
-        )
+        console.log("Temporal components computed.")
+        with console.status("Post-processing components."):
+            self.estimate_components(
+                self.feature_importance,
+            )
+        console.log("Temporal fitting done.")
         return document_topic_matrix
 
     @property