Add fastapitableau-example to Gallery (#77)

Author: fizzboop

.github/workflows/extensions.yml

@@ -38,6 +38,7 @@ jobs:
             landing-page: extensions/landing-page/**
             stock-api-fastapi: extensions/stock-api-fastapi/**
             connectwidgets-example: extensions/connectwidgets-example/**
+            fastapitableau-example: extensions/fastapitableau-example/**
 
   # Runs for each extension that has changed from `simple-extension-changes`
   # Lints and packages in preparation for tests and and release.

extensions/fastapitableau-example/README.md

@@ -0,0 +1,17 @@
+# Tableau Python Extension
+
+## About this example
+
+fastapitableau is a Python package that enables Python developers to build FastAPI APIs that function as Tableau Analytics Extensions. These extensions can be leveraged from Tableau workbooks to allow real time requests from Tableau to Python. This extension builds on top of Tableau's example Superstore dataset.
+
+
+## Learn more
+
+* [Learn how to use this extension from Tableau](https://github.com/sol-eng/tableau-examples/tree/main/superstore)
+* [Documentation for fastapitableau](https://github.com/rstudio/fastapitableau)
+* [Configuring Posit Connect for use with Tableau](https://docs.posit.co/connect/admin/integrations/tableau/)
+
+## Requirements
+
+* Posit Connect license allows API publishing
+* Python version 3.9 or higher

extensions/fastapitableau-example/app.py

@@ -0,0 +1,43 @@
+from fastapitableau import FastAPITableau
+from joblib import load
+import pandas as pd
+from typing import List
+
+# Load model
+model = load("model.joblib")
+
+
+# Define model_pipline
+def model_pipeline(dict):
+    model_data = pd.DataFrame(dict)
+    model_data["ship_diff"] = (
+        model_data.days_to_ship_actual - model_data.days_to_ship_scheduled
+    )
+    pred_columns = ["ship_diff", "quantity", "sales", "discount"]
+    return model.predict(model_data.loc[:, pred_columns]).tolist()
+
+
+# Define the extension
+app = FastAPITableau(
+    title="Predicted Profit",
+    description="A simple linear prediction of sales profit given new input data",
+    version="0.1.0",
+)
+
+
+@app.post("/predict")
+async def predict(
+    days_to_ship_actual: List[int],
+    days_to_ship_scheduled: List[int],
+    quantity: List[int],
+    sales: List[float],
+    discount: List[float],
+) -> List[float]:
+    data = {
+        "days_to_ship_actual": days_to_ship_actual,
+        "days_to_ship_scheduled": days_to_ship_scheduled,
+        "quantity": quantity,
+        "sales": sales,
+        "discount": discount,
+    }
+    return model_pipeline(data)

extensions/fastapitableau-example/clean-orders.csv

@@ -0,0 +1,49 @@
+{
+  "version": 1,
+  "locale": "en_US.UTF-8",
+  "metadata": {
+    "appmode": "python-fastapi",
+    "entrypoint": "app"
+  },
+  "extension": {
+    "name": "fastapitableau-example",
+    "title": "Tableau Python Extension",
+    "description": "fastapitableau is a Python package that enables Python developers to build FastAPI APIs that function as Tableau Analytics Extensions. These extensions can be leveraged from Tableau workbooks to allow real time requests from Tableau to Python. This extension builds on top of Tableau's example Superstore dataset.",
+    "homepage": "https://github.com/posit-dev/connect-extensions/tree/main/extensions/fastapitableau-example",
+    "minimumConnectVersion": "2025.04.0",
+    "version": "1.0.0"
+  },
+  "environment": {
+    "python": {
+      "requires": ">=3.9, <4"
+    }
+  },
+  "python": {
+    "version": "3.11.3",
+    "package_manager": {
+      "name": "pip",
+      "version": "25.0.1",
+      "package_file": "requirements.txt"
+    }
+  },
+  "files": {
+    "requirements.txt": {
+      "checksum": "6a5a141024c9d4ec0a2313e25185e0b2"
+    },
+    "README.md": {
+      "checksum": "6afae511be3a94f9f867276fc3786854"
+    },
+    "app.py": {
+      "checksum": "9c935f4dc153c31b8991ddc8d18ffd7d"
+    },
+    "clean-orders.csv": {
+      "checksum": "f576968ff5b8a02ccc975f45c079d80a"
+    },
+    "model.joblib": {
+      "checksum": "ef047b888ba3b6225f0d55f57b63d466"
+    },
+    "train.py": {
+      "checksum": "c77af4b24184b1e702424c24573444d7"
+    }
+  }
+}

extensions/fastapitableau-example/manifest.json

@@ -0,0 +1,4 @@
+fastapitableau==1.3.0
+joblib==1.3.2
+pandas==2.2.1
+scikit-learn==1.4.1.post1

extensions/fastapitableau-example/model.joblib

@@ -0,0 +1,30 @@
+#
+# Training for fastapitableau-example.
+# Reads clean-orders.csv.
+# Writes model.joblib.
+#
+
+import pandas as pd
+from sklearn.linear_model import LinearRegression
+from joblib import dump
+
+# Read in data
+data = pd.read_csv("clean-orders.csv")
+
+# Define new features and training data
+data["ship_diff"] = data["days_to_ship_actual"] - data["days_to_ship_scheduled"]
+train_columns = [
+    "ship_diff",
+    "quantity",
+    "sales",
+    "discount",
+]
+train_data = data[train_columns]
+
+# Fit linear regression model to predict profit
+reg = LinearRegression().fit(train_data, data.profit)
+reg.predict(train_data)
+data["predicted_profit"] = reg.predict(train_data)
+
+# Write the model for use by the API.
+dump(reg, "model.joblib")