Skip to content

data_processor

Data processing module for the Mangetamain project.

This module contains the :class:DataProcessor class which is responsible for loading raw CSV/ZIP datasets, cleaning and splitting recipe data, merging interactions with recipe metadata and computing simple aggregations used by the frontend (proportions, aggregates and parquet exports).

The class uses Polars for efficient DataFrame operations and writes the processed results into data/processed/ as parquet files.

DataProcessor

Processes and transforms recipe and interaction data.

The :class:DataProcessor is a lightweight ETL helper used to build the processed datasets consumed by the Streamlit frontend. It exposes a small sequence of methods that can be called by a runner or CI job:

  • load_data: Load CSV or extract ZIP and read inputs with Polars.
  • drop_na: Remove rows with missing or unrealistic values.
  • split_minutes: Partition recipes into short/medium/long buckets.
  • merge_data: Join interactions with recipe metadata.
  • compute_proportions: Compute simple aggregates used by plots.
  • save_data: Persist the processed tables to parquet files.
Source code in src/mangetamain/backend/data_processor.py 
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
class DataProcessor:
    """Processes and transforms recipe and interaction data.

    The :class:`DataProcessor` is a lightweight ETL helper used to build
    the processed datasets consumed by the Streamlit frontend. It exposes
    a small sequence of methods that can be called by a runner or CI job:

    - ``load_data``: Load CSV or extract ZIP and read inputs with Polars.
    - ``drop_na``: Remove rows with missing or unrealistic values.
    - ``split_minutes``: Partition recipes into short/medium/long buckets.
    - ``merge_data``: Join interactions with recipe metadata.
    - ``compute_proportions``: Compute simple aggregates used by plots.
    - ``save_data``: Persist the processed tables to parquet files.
    """

    def __init__(
        self,
        data_dir: Path = Path("data/raw"),
        path_interactions: Path = Path("data/raw/RAW_interactions.csv"),
        path_recipes: Path = Path("data/raw/RAW_recipes.csv"),
    ) -> None:
        """Initialize the DataProcessor.

        Args:
            data_dir: Base directory where raw data files live (default ``data/raw``).
            path_interactions: Path to the interactions CSV file.
            path_recipes: Path to the recipes CSV file.
        """
        logger.info("Starting to load data.")
        self.data_dir = data_dir
        self.path_interactions = path_interactions
        self.path_recipes = path_recipes
        self.df_interactions, self.df_recipes = self.load_data()

    def load_data(self) -> tuple[pl.DataFrame, pl.DataFrame]:
        """Load interactions and recipes data.

        The method accepts either CSV files or ZIP archives containing the
        CSVs. If CSV files are not present it will look for ``.csv.zip``
        siblings and extract them into ``data_dir``.

        Returns:
                A tuple ``(df_interactions, df_recipes)`` of Polars DataFrames.
        """
        # Check if CSV files exist, otherwise look for ZIP files
        if not self.path_interactions.exists() or not self.path_recipes.exists():
            logger.info("CSV files not found, checking for ZIP files.")
            path_interaction_zip = self.path_interactions.with_suffix(".csv.zip")
            path_recipes_zip = self.path_recipes.with_suffix(".csv.zip")

            if not path_interaction_zip.exists() or not path_recipes_zip.exists():
                logger.error(
                    f"CSV and ZIP files not found: {path_interaction_zip} and {path_recipes_zip} not found.",
                )
                raise FileNotFoundError(
                    "Neither CSV nor ZIP files found for interactions or recipes.",
                )
            else:
                logger.info("Extracting data from ZIP files.")
                with zipfile.ZipFile(path_interaction_zip, "r") as zip_ref:
                    zip_ref.extractall(self.data_dir)
                with zipfile.ZipFile(path_recipes_zip, "r") as zip_ref:
                    zip_ref.extractall(self.data_dir)

        # Load data from CSV

        try:
            with open(self.path_interactions, "rb") as f:
                df_interactions = pl.read_csv(f, schema_overrides={"date": pl.Datetime})
                logger.info(
                    f"Interactions loaded successfully | Data shape: {df_interactions.shape}.",
                )
            with open(self.path_recipes, "rb") as f:
                df_recipes = pl.read_csv(f, schema_overrides={"submitted": pl.Datetime})
                df_recipes = df_recipes.rename({"id": "recipe_id"})
                logger.info(
                    f"Recipes loaded successfully | Data shape: {df_recipes.shape}.",
                )
        except Exception as e:
            logger.error(f"Error loading CSV files: {e}")
            raise

        return df_interactions, df_recipes

    def drop_na(self) -> None:
        """Drop missing or unrealistic records.

        This method filters out interactions without textual reviews and
        recipes with unrealistic preparation times or zero steps. It
        updates the instance attributes used by downstream processing.
        """
        self.df_interactions_nna = self.df_interactions.filter(
            ~self.df_interactions["review"].is_null(),
        )
        logger.info(
            f"Interactions after dropping NA | Data shape: {self.df_interactions.shape}.",
        )
        self.df_recipes_nna = self.df_recipes.filter(
            (self.df_recipes["minutes"] < 60 * 24 * 365)
            & (self.df_recipes["minutes"] > 0),
        )
        logger.info(
            f"Recipes after dropping unrealistic times | Data shape: {self.df_recipes.shape}.",
        )
        self.df_recipes_nna = self.df_recipes_nna.filter(
            self.df_recipes_nna["n_steps"] > 0,
        )
        logger.info(
            f"Recipes after dropping zero steps | Data shape: {self.df_recipes.shape}.",
        )

    def split_minutes(self) -> None:
        """Split recipes into short, medium, and long buckets based on minutes.

        The thresholds are conservative and chosen to separate quick
        recipes from long projects. Results are stored on the instance as
        ``df_recipes_nna_short``, ``df_recipes_nna_medium`` and
        ``df_recipes_nna_long``.
        """
        self.df_recipes_nna_short = self.df_recipes_nna.filter(
            self.df_recipes_nna["minutes"] <= MEDIUM_LIM,
        )
        self.df_recipes_nna_medium = self.df_recipes_nna.filter(
            (self.df_recipes_nna["minutes"] > MEDIUM_LIM)
            & (self.df_recipes_nna["minutes"] <= LONG_LIM),
        )
        self.df_recipes_nna_long = self.df_recipes_nna.filter(
            self.df_recipes_nna["minutes"] > LONG_LIM,
        )
        logger.info(
            f"Recipes split into short ({self.df_recipes_nna_short.shape}), "
            f"medium ({self.df_recipes_nna_medium.shape}), "
            f"and long ({self.df_recipes_nna_long.shape}).",
        )

    def merge_data(self) -> None:
        """Join interactions with recipes on ``recipe_id``.

        Produces ``total`` tables for each duration bucket that are used to
        compute rating proportions and other aggregates.
        """
        self.total_nt = self.df_interactions.join(
            self.df_recipes,
            on="recipe_id",
            how="inner",
        )
        self.total = self.df_interactions_nna.join(
            self.df_recipes_nna,
            on="recipe_id",
            how="inner",
        )
        self.total_short = self.df_interactions_nna.join(
            self.df_recipes_nna_short,
            on="recipe_id",
            how="inner",
        )
        self.total_medium = self.df_interactions_nna.join(
            self.df_recipes_nna_medium,
            on="recipe_id",
            how="inner",
        )
        self.total_long = self.df_interactions_nna.join(
            self.df_recipes_nna_long,
            on="recipe_id",
            how="inner",
        )
        logger.info(f"Merged data shape: {self.total.shape}.")
        logger.info(f"Merged short recipes data shape: {self.total_short.shape}.")
        logger.info(f"Merged medium recipes data shape: {self.total_medium.shape}.")
        logger.info(f"Merged long recipes data shape: {self.total_long.shape}.")

    def compute_proportions(self) -> None:
        """Compute 5-star proportions by preparation time and number of steps.

        This method fills ``df_proportion_m`` and ``df_proportion_s`` which
        contain the proportion of 5-star ratings aggregated by minute and
        number-of-steps respectively. Results are suitable for plotting in
        the frontend.
        """
        # minutes = np.array(sorted(self.df_recipes_nna_court["minutes"].unique()))
        logger.info("Computing proportions of 5-star ratings by minutes")
        minutes = np.array(sorted(self.df_recipes_nna_short["minutes"].unique()))
        comptes = (
            self.total_short["minutes"]
            .value_counts()
            .sort("minutes")["count"]
            .to_numpy()
        )
        proportions = (
            self.total_short.filter(pl.col("rating") == RATING_MAX)["minutes"]
            .value_counts()
            .sort("minutes")["count"]
            .to_numpy()
        )
        proportion_m = proportions / comptes

        # steps = np.array(sorted(self.df_recipes_nna[self.df_recipes_nna["n_steps"] <= 40].n_steps.unique()))
        logger.info("Computing proportions of 5-star ratings by steps")
        steps = np.array(
            sorted(
                self.df_recipes_nna.filter(pl.col("n_steps") <= NB_STEPS_MAX)[
                    "n_steps"
                ].unique(),
            ),
        )
        comptes = (
            self.total.filter(pl.col("n_steps") <= NB_STEPS_MAX)["n_steps"]
            .value_counts()
            .sort("n_steps")["count"]
            .to_numpy()
        )
        proportions = (
            self.total.filter(
                (pl.col("n_steps") <= NB_STEPS_MAX) & (pl.col("rating") == RATING_MAX),
            )["n_steps"]
            .value_counts()
            .sort("n_steps")["count"]
            .to_numpy()
        )
        proportion_s = proportions / comptes

        logger.info("Proportions computed. Loading internally")
        self.df_proportion_m = pl.DataFrame(
            {
                "minutes": minutes.astype(int),
                "proportion_m": proportion_m.astype(float),
            },
        )  # type conversion needed for parquet
        self.df_proportion_s = pl.DataFrame(
            {"n_steps": steps.astype(int), "proportion_s": proportion_s.astype(float)},
        )

    def process_recipes(self) -> None:
        """Create a RecipeAnalyzer instance for NLP and visualization.

        Initializes a :class:`RecipeAnalyzer` with the loaded data and stores
        it as ``self.recipe_analyzer``. This object provides word cloud generation,
        TF-IDF analysis, and other recipe text analysis features.
        """
        self.recipe_analyzer = RecipeAnalyzer(
            self.df_interactions,
            self.df_recipes,
            self.total,
        )

    def user_df(self) -> None:
        """Compute user-level aggregates and store as df_user."""
        logger.info("Computing user-level aggregates")
        self.df_user = self.total.group_by("user_id").agg(
            nb_reviews=pl.len(),
            mean_rating=pl.col("rating").mean(),
            std_rating=pl.col("rating").std(),
            review_length=pl.col("review").flatten().str.split(" ").list.len().mean(),
            mean_time=pl.col("minutes").mean(),
        )

        self.df_user = self.df_user.with_columns(
            pl.col("std_rating").fill_nan(0).fill_null(0).alias("std_rating")
        )

    def save_data(self) -> None:
        """Persist processed tables to parquet files under ``data/processed/``.

        The output files are:
        - ``processed_interactions.parquet``
        - ``processed_recipes.parquet``
        - ``total.parquet`` (merged interactions)
        - ``short.parquet`` (merged short recipes)
        - ``proportion_m.parquet`` and ``proportion_s.parquet``
        """
        logger.info("Starting to save the data in parquet")
        save_folder = Path("data/processed")
        save_folder.mkdir(parents=True, exist_ok=True)
        logger.info("Saving df_interactions")
        self.df_interactions.write_parquet(
            "data/processed/initial_interactions.parquet",
        )
        self.df_interactions_nna.write_parquet(
            "data/processed/processed_interactions.parquet",
        )
        logger.info("Done \n Saving df_recipes")
        self.df_recipes.write_parquet("data/processed/initial_recipes.parquet")
        self.df_recipes_nna.write_parquet("data/processed/processed_recipes.parquet")

        logger.info("Done \n Saving total data")
        self.total_nt.write_parquet("data/processed/total_nt.parquet")
        self.total.write_parquet("data/processed/total.parquet")

        logger.info("Done \n Saving total short data")
        self.total_short.write_parquet("data/processed/short.parquet")
        # self.df_recipes_nna_medium.write_parquet("data/processed/medium.parquet")
        # self.df_recipes_nna_long.write_parquet("data/processed/long.parquet")

        logger.info("Done \n Saving proportions data")
        self.df_proportion_m.write_parquet("data/processed/proportion_m.parquet")
        self.df_proportion_s.write_parquet("data/processed/proportion_s.parquet")

        logger.info("Done \n Saving recipe analyzer object")

        self.recipe_analyzer.save("data/processed/recipe_analyzer.pkl")

        logger.info("Done \n Saving user data")

        self.df_user.write_parquet("data/processed/user.parquet")

        logger.info("All processed data saved to parquet files.")

__init__ 

__init__(
    data_dir=Path("data/raw"),
    path_interactions=Path("data/raw/RAW_interactions.csv"),
    path_recipes=Path("data/raw/RAW_recipes.csv"),
)

Initialize the DataProcessor.

Parameters:

Name Type Description Default
data_dir Path

Base directory where raw data files live (default data/raw).

 Path('data/raw')
path_interactions Path

Path to the interactions CSV file.

 Path('data/raw/RAW_interactions.csv')
path_recipes Path

Path to the recipes CSV file.

 Path('data/raw/RAW_recipes.csv')
Source code in src/mangetamain/backend/data_processor.py 
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
def __init__(
    self,
    data_dir: Path = Path("data/raw"),
    path_interactions: Path = Path("data/raw/RAW_interactions.csv"),
    path_recipes: Path = Path("data/raw/RAW_recipes.csv"),
) -> None:
    """Initialize the DataProcessor.

    Args:
        data_dir: Base directory where raw data files live (default ``data/raw``).
        path_interactions: Path to the interactions CSV file.
        path_recipes: Path to the recipes CSV file.
    """
    logger.info("Starting to load data.")
    self.data_dir = data_dir
    self.path_interactions = path_interactions
    self.path_recipes = path_recipes
    self.df_interactions, self.df_recipes = self.load_data()

compute_proportions 

compute_proportions()

Compute 5-star proportions by preparation time and number of steps.

This method fills df_proportion_m and df_proportion_s which contain the proportion of 5-star ratings aggregated by minute and number-of-steps respectively. Results are suitable for plotting in the frontend.

Source code in src/mangetamain/backend/data_processor.py 
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
def compute_proportions(self) -> None:
    """Compute 5-star proportions by preparation time and number of steps.

    This method fills ``df_proportion_m`` and ``df_proportion_s`` which
    contain the proportion of 5-star ratings aggregated by minute and
    number-of-steps respectively. Results are suitable for plotting in
    the frontend.
    """
    # minutes = np.array(sorted(self.df_recipes_nna_court["minutes"].unique()))
    logger.info("Computing proportions of 5-star ratings by minutes")
    minutes = np.array(sorted(self.df_recipes_nna_short["minutes"].unique()))
    comptes = (
        self.total_short["minutes"]
        .value_counts()
        .sort("minutes")["count"]
        .to_numpy()
    )
    proportions = (
        self.total_short.filter(pl.col("rating") == RATING_MAX)["minutes"]
        .value_counts()
        .sort("minutes")["count"]
        .to_numpy()
    )
    proportion_m = proportions / comptes

    # steps = np.array(sorted(self.df_recipes_nna[self.df_recipes_nna["n_steps"] <= 40].n_steps.unique()))
    logger.info("Computing proportions of 5-star ratings by steps")
    steps = np.array(
        sorted(
            self.df_recipes_nna.filter(pl.col("n_steps") <= NB_STEPS_MAX)[
                "n_steps"
            ].unique(),
        ),
    )
    comptes = (
        self.total.filter(pl.col("n_steps") <= NB_STEPS_MAX)["n_steps"]
        .value_counts()
        .sort("n_steps")["count"]
        .to_numpy()
    )
    proportions = (
        self.total.filter(
            (pl.col("n_steps") <= NB_STEPS_MAX) & (pl.col("rating") == RATING_MAX),
        )["n_steps"]
        .value_counts()
        .sort("n_steps")["count"]
        .to_numpy()
    )
    proportion_s = proportions / comptes

    logger.info("Proportions computed. Loading internally")
    self.df_proportion_m = pl.DataFrame(
        {
            "minutes": minutes.astype(int),
            "proportion_m": proportion_m.astype(float),
        },
    )  # type conversion needed for parquet
    self.df_proportion_s = pl.DataFrame(
        {"n_steps": steps.astype(int), "proportion_s": proportion_s.astype(float)},
    )

drop_na 

drop_na()

Drop missing or unrealistic records.

This method filters out interactions without textual reviews and recipes with unrealistic preparation times or zero steps. It updates the instance attributes used by downstream processing.

Source code in src/mangetamain/backend/data_processor.py 
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
def drop_na(self) -> None:
    """Drop missing or unrealistic records.

    This method filters out interactions without textual reviews and
    recipes with unrealistic preparation times or zero steps. It
    updates the instance attributes used by downstream processing.
    """
    self.df_interactions_nna = self.df_interactions.filter(
        ~self.df_interactions["review"].is_null(),
    )
    logger.info(
        f"Interactions after dropping NA | Data shape: {self.df_interactions.shape}.",
    )
    self.df_recipes_nna = self.df_recipes.filter(
        (self.df_recipes["minutes"] < 60 * 24 * 365)
        & (self.df_recipes["minutes"] > 0),
    )
    logger.info(
        f"Recipes after dropping unrealistic times | Data shape: {self.df_recipes.shape}.",
    )
    self.df_recipes_nna = self.df_recipes_nna.filter(
        self.df_recipes_nna["n_steps"] > 0,
    )
    logger.info(
        f"Recipes after dropping zero steps | Data shape: {self.df_recipes.shape}.",
    )

load_data 

load_data()

Load interactions and recipes data.

The method accepts either CSV files or ZIP archives containing the CSVs. If CSV files are not present it will look for .csv.zip siblings and extract them into data_dir.

Returns:

Type Description
tuple[DataFrame, DataFrame]

A tuple (df_interactions, df_recipes) of Polars DataFrames.
Source code in src/mangetamain/backend/data_processor.py 
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
def load_data(self) -> tuple[pl.DataFrame, pl.DataFrame]:
    """Load interactions and recipes data.

    The method accepts either CSV files or ZIP archives containing the
    CSVs. If CSV files are not present it will look for ``.csv.zip``
    siblings and extract them into ``data_dir``.

    Returns:
            A tuple ``(df_interactions, df_recipes)`` of Polars DataFrames.
    """
    # Check if CSV files exist, otherwise look for ZIP files
    if not self.path_interactions.exists() or not self.path_recipes.exists():
        logger.info("CSV files not found, checking for ZIP files.")
        path_interaction_zip = self.path_interactions.with_suffix(".csv.zip")
        path_recipes_zip = self.path_recipes.with_suffix(".csv.zip")

        if not path_interaction_zip.exists() or not path_recipes_zip.exists():
            logger.error(
                f"CSV and ZIP files not found: {path_interaction_zip} and {path_recipes_zip} not found.",
            )
            raise FileNotFoundError(
                "Neither CSV nor ZIP files found for interactions or recipes.",
            )
        else:
            logger.info("Extracting data from ZIP files.")
            with zipfile.ZipFile(path_interaction_zip, "r") as zip_ref:
                zip_ref.extractall(self.data_dir)
            with zipfile.ZipFile(path_recipes_zip, "r") as zip_ref:
                zip_ref.extractall(self.data_dir)

    # Load data from CSV

    try:
        with open(self.path_interactions, "rb") as f:
            df_interactions = pl.read_csv(f, schema_overrides={"date": pl.Datetime})
            logger.info(
                f"Interactions loaded successfully | Data shape: {df_interactions.shape}.",
            )
        with open(self.path_recipes, "rb") as f:
            df_recipes = pl.read_csv(f, schema_overrides={"submitted": pl.Datetime})
            df_recipes = df_recipes.rename({"id": "recipe_id"})
            logger.info(
                f"Recipes loaded successfully | Data shape: {df_recipes.shape}.",
            )
    except Exception as e:
        logger.error(f"Error loading CSV files: {e}")
        raise

    return df_interactions, df_recipes

merge_data 

merge_data()

Join interactions with recipes on recipe_id.

Produces total tables for each duration bucket that are used to compute rating proportions and other aggregates.

Source code in src/mangetamain/backend/data_processor.py 
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
def merge_data(self) -> None:
    """Join interactions with recipes on ``recipe_id``.

    Produces ``total`` tables for each duration bucket that are used to
    compute rating proportions and other aggregates.
    """
    self.total_nt = self.df_interactions.join(
        self.df_recipes,
        on="recipe_id",
        how="inner",
    )
    self.total = self.df_interactions_nna.join(
        self.df_recipes_nna,
        on="recipe_id",
        how="inner",
    )
    self.total_short = self.df_interactions_nna.join(
        self.df_recipes_nna_short,
        on="recipe_id",
        how="inner",
    )
    self.total_medium = self.df_interactions_nna.join(
        self.df_recipes_nna_medium,
        on="recipe_id",
        how="inner",
    )
    self.total_long = self.df_interactions_nna.join(
        self.df_recipes_nna_long,
        on="recipe_id",
        how="inner",
    )
    logger.info(f"Merged data shape: {self.total.shape}.")
    logger.info(f"Merged short recipes data shape: {self.total_short.shape}.")
    logger.info(f"Merged medium recipes data shape: {self.total_medium.shape}.")
    logger.info(f"Merged long recipes data shape: {self.total_long.shape}.")

process_recipes 

process_recipes()

Create a RecipeAnalyzer instance for NLP and visualization.

Initializes a :class:RecipeAnalyzer with the loaded data and stores it as self.recipe_analyzer. This object provides word cloud generation, TF-IDF analysis, and other recipe text analysis features.

Source code in src/mangetamain/backend/data_processor.py 
262
263
264
265
266
267
268
269
270
271
272
273
def process_recipes(self) -> None:
    """Create a RecipeAnalyzer instance for NLP and visualization.

    Initializes a :class:`RecipeAnalyzer` with the loaded data and stores
    it as ``self.recipe_analyzer``. This object provides word cloud generation,
    TF-IDF analysis, and other recipe text analysis features.
    """
    self.recipe_analyzer = RecipeAnalyzer(
        self.df_interactions,
        self.df_recipes,
        self.total,
    )

save_data 

save_data()

Persist processed tables to parquet files under data/processed/.

The output files are: - processed_interactions.parquet - processed_recipes.parquet - total.parquet (merged interactions) - short.parquet (merged short recipes) - proportion_m.parquet and proportion_s.parquet

Source code in src/mangetamain/backend/data_processor.py 
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
def save_data(self) -> None:
    """Persist processed tables to parquet files under ``data/processed/``.

    The output files are:
    - ``processed_interactions.parquet``
    - ``processed_recipes.parquet``
    - ``total.parquet`` (merged interactions)
    - ``short.parquet`` (merged short recipes)
    - ``proportion_m.parquet`` and ``proportion_s.parquet``
    """
    logger.info("Starting to save the data in parquet")
    save_folder = Path("data/processed")
    save_folder.mkdir(parents=True, exist_ok=True)
    logger.info("Saving df_interactions")
    self.df_interactions.write_parquet(
        "data/processed/initial_interactions.parquet",
    )
    self.df_interactions_nna.write_parquet(
        "data/processed/processed_interactions.parquet",
    )
    logger.info("Done \n Saving df_recipes")
    self.df_recipes.write_parquet("data/processed/initial_recipes.parquet")
    self.df_recipes_nna.write_parquet("data/processed/processed_recipes.parquet")

    logger.info("Done \n Saving total data")
    self.total_nt.write_parquet("data/processed/total_nt.parquet")
    self.total.write_parquet("data/processed/total.parquet")

    logger.info("Done \n Saving total short data")
    self.total_short.write_parquet("data/processed/short.parquet")
    # self.df_recipes_nna_medium.write_parquet("data/processed/medium.parquet")
    # self.df_recipes_nna_long.write_parquet("data/processed/long.parquet")

    logger.info("Done \n Saving proportions data")
    self.df_proportion_m.write_parquet("data/processed/proportion_m.parquet")
    self.df_proportion_s.write_parquet("data/processed/proportion_s.parquet")

    logger.info("Done \n Saving recipe analyzer object")

    self.recipe_analyzer.save("data/processed/recipe_analyzer.pkl")

    logger.info("Done \n Saving user data")

    self.df_user.write_parquet("data/processed/user.parquet")

    logger.info("All processed data saved to parquet files.")

split_minutes 

split_minutes()

Split recipes into short, medium, and long buckets based on minutes.

The thresholds are conservative and chosen to separate quick recipes from long projects. Results are stored on the instance as df_recipes_nna_short, df_recipes_nna_medium and df_recipes_nna_long.

Source code in src/mangetamain/backend/data_processor.py 
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
def split_minutes(self) -> None:
    """Split recipes into short, medium, and long buckets based on minutes.

    The thresholds are conservative and chosen to separate quick
    recipes from long projects. Results are stored on the instance as
    ``df_recipes_nna_short``, ``df_recipes_nna_medium`` and
    ``df_recipes_nna_long``.
    """
    self.df_recipes_nna_short = self.df_recipes_nna.filter(
        self.df_recipes_nna["minutes"] <= MEDIUM_LIM,
    )
    self.df_recipes_nna_medium = self.df_recipes_nna.filter(
        (self.df_recipes_nna["minutes"] > MEDIUM_LIM)
        & (self.df_recipes_nna["minutes"] <= LONG_LIM),
    )
    self.df_recipes_nna_long = self.df_recipes_nna.filter(
        self.df_recipes_nna["minutes"] > LONG_LIM,
    )
    logger.info(
        f"Recipes split into short ({self.df_recipes_nna_short.shape}), "
        f"medium ({self.df_recipes_nna_medium.shape}), "
        f"and long ({self.df_recipes_nna_long.shape}).",
    )

user_df 

user_df()

Compute user-level aggregates and store as df_user.

Source code in src/mangetamain/backend/data_processor.py 
275
276
277
278
279
280
281
282
283
284
285
286
287
288
def user_df(self) -> None:
    """Compute user-level aggregates and store as df_user."""
    logger.info("Computing user-level aggregates")
    self.df_user = self.total.group_by("user_id").agg(
        nb_reviews=pl.len(),
        mean_rating=pl.col("rating").mean(),
        std_rating=pl.col("rating").std(),
        review_length=pl.col("review").flatten().str.split(" ").list.len().mean(),
        mean_time=pl.col("minutes").mean(),
    )

    self.df_user = self.df_user.with_columns(
        pl.col("std_rating").fill_nan(0).fill_null(0).alias("std_rating")
    )