poolish-ds-data 0.1.0__py3-none-any.whl

poolish_ds_data/__init__.py

@@ -0,0 +1,66 @@
+import pandas as pd
+import requests
+from io import StringIO
+
+BASE_URL = "https://raw.githubusercontent.com/sawula/ds-course-data/main/"
+
+DATASETS = [
+    "FC26_20250921.csv",
+    "Madden25.csv",
+    "MER_T02_01A.csv",
+    "NBA Stats 2324.csv",
+    "NFLsalaries.csv",
+    "NYC ER respire 16 20 totals.csv",
+    "NYC ER respire 16 20.csv",
+    "PandasJumble0.csv",
+    "PandasJumble1.csv",
+    "PandasJumble2.csv",
+    "SF2506_stations_with_elevation.csv",
+    "WS_golf_winnings.csv",
+    "brainrot.csv",
+    "coldmed.csv",
+    "combine_stat_crunch.csv",
+    "data8station.csv",
+    "data8trip.csv",
+    "dc_v_state.csv",
+    "electric_price_state_sector_23_24.csv",
+    "electric_prices_state_sector_23_24.csv",
+    "epa_02_10.csv",
+    "epl2122.csv",
+    "pandemic_trends.csv",
+    "players_22.csv",
+    "pop_shift_0.csv",
+    "resume.csv",
+    "resumes_city_eoe.csv",
+    "state_pop.csv",
+    "top_colleges.csv",
+]
+
+
+def load(filename):
+    """Load a dataset by filename. Returns a pandas DataFrame for CSV files,
+    or a dict for JSON/GeoJSON files."""
+    url = BASE_URL + filename.replace(" ", "%20")
+    response = requests.get(url)
+    response.raise_for_status()
+
+    if filename.endswith(".csv"):
+        return pd.read_csv(StringIO(response.text))
+    elif filename.endswith(".json") or filename.endswith(".geojson"):
+        return response.json()
+    else:
+        raise ValueError(
+            f"Unsupported file type for direct loading. "
+            f"Use get_url('{filename}') to get the raw URL instead."
+        )
+
+
+def list_datasets():
+    """Print all available datasets."""
+    for name in sorted(DATASETS):
+        print(name)
+
+
+def get_url(filename):
+    """Get the raw GitHub URL for any file in the repo."""
+    return BASE_URL + filename.replace(" ", "%20")

poolish_ds_data-0.1.0.dist-info/METADATA

@@ -0,0 +1,37 @@
+Metadata-Version: 2.4
+Name: poolish-ds-data
+Version: 0.1.0
+Summary: Data files for Peddie School data science course
+Author-email: Mark Sawula <msawula@peddie.org>
+License: MIT
+Project-URL: Homepage, https://sawula.github.io
+Project-URL: Repository, https://github.com/sawula/ds-course-data
+Requires-Python: >=3.8
+Description-Content-Type: text/markdown
+Requires-Dist: pandas
+Requires-Dist: requests
+
+# poolish-ds-data
+
+Data files for the Peddie School data science course.
+
+## Installation
+
+```bash
+pip install poolish-ds-data
+```
+
+## Usage
+
+```python
+import poolish_ds_data as ds
+
+# Load a CSV as a pandas DataFrame
+df = ds.load("epl2122.csv")
+
+# See all available datasets
+ds.list_datasets()
+
+# Get the raw URL for any file (useful for images, GeoJSON, etc.)
+url = ds.get_url("south-korea-with-regions_1516.geojson")
+```

poolish_ds_data-0.1.0.dist-info/RECORD

@@ -0,0 +1,5 @@
+poolish_ds_data/__init__.py,sha256=5y7zAiz3xRRw0JGeyTquH_ISXlnBC8Ze_90N4FHippc,1784
+poolish_ds_data-0.1.0.dist-info/METADATA,sha256=j4kkGGaDaD7jIIa7R2kArO8KMXO5Axy8czM5UX_0pO0,824
+poolish_ds_data-0.1.0.dist-info/WHEEL,sha256=aeYiig01lYGDzBgS8HxWXOg3uV61G9ijOsup-k9o1sk,91
+poolish_ds_data-0.1.0.dist-info/top_level.txt,sha256=9l2FolHaj57zh5PmSUK89j9ACoCp3zfCsxxMM0eAgU8,16
+poolish_ds_data-0.1.0.dist-info/RECORD,,

poolish_ds_data-0.1.0.dist-info/WHEEL

@@ -0,0 +1,5 @@
+Wheel-Version: 1.0
+Generator: setuptools (82.0.1)
+Root-Is-Purelib: true
+Tag: py3-none-any
+

poolish_ds_data-0.1.0.dist-info/top_level.txt

@@ -0,0 +1 @@
+poolish_ds_data