Adding the libre textbooks

data/libre_textbooks/meta.yaml

@@ -0,0 +1,30 @@
+---
+name: libre_textbooks
+description: A dataset of scraped articles from libre textbooks
+targets:
+    - id: html
+      description: A scraped page from libre textbooks
+      units:
+      type: text
+      names:
+          - natural language article
+      pubchem_aids: []
+      uris: []
+identifiers:
+    - id: url 
+      type: string
+      description: url of the page the content is scraped from
+    - id: text_length
+      type: int
+      description: text character count
+license: CC BY 4.0
+links:
+    - name: Libre Textbooks
+      url: https://chem.libretexts.org/Bookshelves
+      description: ''
+    - name: Hugging Face dataset upload
+      url: https://huggingface.co/datasets/Hack90/libre_chem_textbooks
+      description: Hugging Face dataset uploaded to HF account
+benchmarks: []
+num_points: 3740
+bibtex: []

data/libre_textbooks/transform.py

@@ -0,0 +1,76 @@
+from datasets import load_dataset
+import pandas as pd
+import yaml
+
+
+LINES_TO_REMOVE = "/workspaces/chemnlp/data/libre_textbooks/lines_to_remove.jsonl"
+RAW_DATASET = "Hack90/libre_chem_textbooks"
+
+
+META_YAML_PATH = "./data/libre_textbooks/meta.yaml"
+META_TEMPLATE = {
+    "name": "libre_textbooks",  # unique identifier, we will also use this for directory names
+    "description": "A dataset of scraped articles from libre textbooks",
+    "targets": [
+        {
+            "id": "html",  # name of the column in a tabular dataset
+            "description": "A scraped page from libre textbooks",
+            "units": None,  # units of the values in this column (leave empty if unitless)
+            "type": "string",  # can be "categorical", "ordinal", "continuous", "string"
+            "names": [  # names for the property (to sample from for building the prompts)
+                "natural language article",
+            ],
+            "pubchem_aids": [],
+            "uris": [],
+        },
+    ],
+    "identifiers": [
+        {
+            "id": "url",  # column name
+            "type": "Other",  # can be "SMILES", "SELFIES", "IUPAC", "OTHER"
+            "description": "url of the page the content is scraped from",
+        },
+    ],
+    "license": "CC BY 4.0",  # license under which the original dataset was published
+    "links": [  # list of relevant links (original dataset, other uses, etc.)
+        {
+            "name": "Libre Textbooks",
+            "url": "https://chem.libretexts.org/Bookshelves",
+            "description": "",
+        },
+        {
+            "name": "Hugging Face dataset upload",
+            "url": "https://huggingface.co/datasets/Hack90/libre_chem_textbooks",
+            "description": "Hugging Face dataset uploaded to HF account",  # Hopefully will move this
+            # to the openbioml space
+        },
+    ],
+    "benchmarks": [],
+    "num_points": 3740,  # number of datapoints in this dataset
+    "bibtex": [
+        # noqa
+    ],
+}
+
+
+def get_raw_data(raw_dataset: str = RAW_DATASET) -> pd.DataFrame:
+    """Load the raw dataset into a pandas dataframe"""
+    dataset = load_dataset(raw_dataset)
+    df_raw = pd.DataFrame(dataset["train"].to_pandas())
+    return df_raw
+
+
+def create_meta_yaml(num_points: int):
+    """Create meta configuration file for the dataset"""
+    # create meta yaml
+    META_TEMPLATE["num_points"] = num_points
+    with open(META_YAML_PATH, "w+") as f:
+        yaml.dump(META_TEMPLATE, f, sort_keys=False)
+    print(f"Finished processing libre_textbooks {META_TEMPLATE['name']} dataset!")
+
+
+if __name__ == "__main__":
+    num_samples = 0
+    raw_df = get_raw_data()
+    num_samples += len(raw_df)
+    create_meta_yaml(num_samples)