From c31743f0669c12290232afefa6ef6de0380b21c5 Mon Sep 17 00:00:00 2001
From: Stefano Pigozzi <me@steffo.eu>
Date: Tue, 7 Feb 2023 10:22:09 +0100
Subject: [PATCH] back to i have no idea of what's happening, but at least it
 works

---
 .idea/runConfigurations/unimore_bda_6.xml |  3 +-
 unimore_bda_6/analysis/base.py            |  3 +-
 unimore_bda_6/analysis/tf_text.py         | 45 ++++++++++++++++-------
 unimore_bda_6/database/cache.py           |  1 -
 unimore_bda_6/database/datatypes.py       |  8 +++-
 unimore_bda_6/log.py                      |  2 +-
 6 files changed, 42 insertions(+), 20 deletions(-)

diff --git a/.idea/runConfigurations/unimore_bda_6.xml b/.idea/runConfigurations/unimore_bda_6.xml
index 77fdf31..fa5fcae 100644
--- a/.idea/runConfigurations/unimore_bda_6.xml
+++ b/.idea/runConfigurations/unimore_bda_6.xml
@@ -5,10 +5,11 @@
     <option name="PARENT_ENVS" value="true" />
     <envs>
       <env name="CONFIRM_OVERWRITE" value="False" />
-      <env name="DATA_SET_SIZE" value="100" />
+      <env name="DATA_SET_SIZE" value="2500" />
       <env name="NLTK_DATA" value="./data/nltk" />
       <env name="PYTHONUNBUFFERED" value="1" />
       <env name="WORKING_SET_SIZE" value="1000000" />
+      <env name="XLA_FLAGS" value="--xla_gpu_cuda_data_dir=/opt/cuda" />
     </envs>
     <option name="SDK_HOME" value="$PROJECT_DIR$/.venv/bin/python" />
     <option name="SDK_NAME" value="Poetry (unimore-bda-6)" />
diff --git a/unimore_bda_6/analysis/base.py b/unimore_bda_6/analysis/base.py
index b9c3900..eea07b4 100644
--- a/unimore_bda_6/analysis/base.py
+++ b/unimore_bda_6/analysis/base.py
@@ -1,9 +1,8 @@
 import abc
 import logging
-import typing as t
 import dataclasses
 
-from ..database import Text, Category, Review, DatasetFunc
+from ..database import Text, Category, DatasetFunc
 
 log = logging.getLogger(__name__)
 
diff --git a/unimore_bda_6/analysis/tf_text.py b/unimore_bda_6/analysis/tf_text.py
index 7ac3825..8beee6d 100644
--- a/unimore_bda_6/analysis/tf_text.py
+++ b/unimore_bda_6/analysis/tf_text.py
@@ -1,6 +1,7 @@
 import tensorflow
 
 from ..database import Text, Category, DatasetFunc
+from ..config import DATA_SET_SIZE
 from .base import BaseSentimentAnalyzer, AlreadyTrainedError, NotTrainedError
 
 
@@ -21,17 +22,20 @@ class TensorflowSentimentAnalyzer(BaseSentimentAnalyzer):
             dataset_func_with_tensor_tuple,
             output_signature=(
                 tensorflow.TensorSpec(shape=(), dtype=tensorflow.string, name="text"),
-                tensorflow.TensorSpec(shape=(), dtype=tensorflow.float32, name="category"),
+                tensorflow.TensorSpec(shape=(5,), dtype=tensorflow.float32, name="category"),
             )
         )
 
     def _build_model(self) -> tensorflow.keras.Sequential:
         return tensorflow.keras.Sequential([
-            tensorflow.keras.layers.Embedding(input_dim=self.MAX_FEATURES + 1, output_dim=self.EMBEDDING_DIM),
-            tensorflow.keras.layers.Dropout(0.2),
+            tensorflow.keras.layers.Embedding(
+                input_dim=self.MAX_FEATURES + 1,
+                output_dim=self.EMBEDDING_DIM,
+            ),
+            # tensorflow.keras.layers.Dropout(0.2),
             tensorflow.keras.layers.GlobalAveragePooling1D(),
-            tensorflow.keras.layers.Dropout(0.2),
-            tensorflow.keras.layers.Dense(1),
+            # tensorflow.keras.layers.Dropout(0.2),
+            tensorflow.keras.layers.Dense(5, activation="softmax"),
         ])
 
     def _build_vectorizer(self) -> tensorflow.keras.layers.TextVectorization:
@@ -41,9 +45,13 @@ class TensorflowSentimentAnalyzer(BaseSentimentAnalyzer):
         text = tensorflow.expand_dims(text, -1)  # TODO: ??????
         return self.text_vectorization_layer(text), category
 
-    MAX_FEATURES = 1000
-    EMBEDDING_DIM = 16
-    EPOCHS = 10
+    MAX_FEATURES = 2500
+    EMBEDDING_DIM = 24
+    """
+    Count of possible "semantic meanings" of words, represented as dimensions of a tensor.
+    """
+
+    EPOCHS = 3
 
     def train(self, dataset_func: DatasetFunc) -> None:
         if self.trained:
@@ -55,11 +63,10 @@ class TensorflowSentimentAnalyzer(BaseSentimentAnalyzer):
         self.text_vectorization_layer.adapt(only_text_set)
         training_set = training_set.map(self.__vectorize_data)
 
-        self.model.compile(loss=tensorflow.keras.losses.CosineSimilarity(axis=0), metrics=["accuracy"])
+        # self.model.compile(loss=tensorflow.keras.losses.SparseCategoricalCrossentropy(from_logits=True), optimizer="adam", metrics=["accuracy"])
+        self.model.compile(loss=tensorflow.keras.losses.MeanAbsoluteError(), optimizer="adam", metrics=["accuracy"])
 
-        history = self.model.fit(training_set, epochs=self.EPOCHS)
-
-        ...
+        self.model.fit(training_set, epochs=self.EPOCHS)
 
         self.trained = True
 
@@ -67,5 +74,15 @@ class TensorflowSentimentAnalyzer(BaseSentimentAnalyzer):
         if not self.trained:
             raise NotTrainedError()
 
-        prediction = self.model.predict(text)
-        breakpoint()
+        vector = self.text_vectorization_layer(tensorflow.expand_dims(text, -1))
+
+        prediction = self.model.predict(vector)
+
+        max_i = None
+        max_p = None
+        for i, p in enumerate(iter(prediction[0])):
+            if max_p is None or p > max_p:
+                max_i = i
+                max_p = p
+
+        return float(max_i) + 1.0
diff --git a/unimore_bda_6/database/cache.py b/unimore_bda_6/database/cache.py
index 9db9829..ae00f34 100644
--- a/unimore_bda_6/database/cache.py
+++ b/unimore_bda_6/database/cache.py
@@ -41,7 +41,6 @@ def load_cache(path: str | pathlib.Path) -> DatasetFunc:
     path = pathlib.Path(path)
 
     if not path.exists():
-        log.error("Specified cache directory does not exist: %s", path)
         raise FileNotFoundError("The specified path does not exist.")
 
     def data_cache_loader():
diff --git a/unimore_bda_6/database/datatypes.py b/unimore_bda_6/database/datatypes.py
index 32c65f1..5f1df35 100644
--- a/unimore_bda_6/database/datatypes.py
+++ b/unimore_bda_6/database/datatypes.py
@@ -33,7 +33,13 @@ class Review:
         return tensorflow.convert_to_tensor(self.text, dtype=tensorflow.string)
 
     def to_tensor_category(self) -> tensorflow.Tensor:
-        return tensorflow.convert_to_tensor(self.category / 5.0, dtype=tensorflow.float32)
+        return tensorflow.convert_to_tensor([
+            1.0 if self.category == 1.0 else 0.0,
+            1.0 if self.category == 2.0 else 0.0,
+            1.0 if self.category == 3.0 else 0.0,
+            1.0 if self.category == 4.0 else 0.0,
+            1.0 if self.category == 5.0 else 0.0,
+        ], dtype=tensorflow.float32)
 
     def to_tensor_tuple(self) -> tuple[tensorflow.Tensor, tensorflow.Tensor]:
         return (
diff --git a/unimore_bda_6/log.py b/unimore_bda_6/log.py
index 4789be8..c7272fe 100644
--- a/unimore_bda_6/log.py
+++ b/unimore_bda_6/log.py
@@ -15,7 +15,7 @@ def install_log_handler(loggers: list[logging.Logger] = None):
     for logger in loggers:
         coloredlogs.install(
             logger=logger,
-            level="DEBUG",
+            level="INFO",
             fmt="{asctime} | {name:<32} | {levelname:>8} | {message}",
             style="{",
             level_styles=dict(