Halyna R.

AI_ASSIST.md

@@ -7,28 +7,41 @@ Document one place you used an LLM during this assignment.
 <!-- TODO: describe the specific problem you asked an LLM about.
      Example: "My Streamlit KPI panel kept re-querying Postgres on every
      sidebar interaction even though I wrapped run_query in @st.cache_data." -->
+While working on the Week 11 Streamlit dashboard, I got a PostgreSQL error when the app tried to query `dev_halyna.fct_trips`.
 
-TODO
+The error was:
+
+`relation "dev_halyna.fct_trips" does not exist`
+
+At first I thought the problem was in my `SELECT` statement, but I was not sure whether the issue came from the SQL query, the database schema, or the Streamlit `.env` configuration.
 
 ## The prompt
 
 <!-- TODO: paste the exact prompt you sent to the LLM. -->
 
-TODO
+I asked the LLM:
+
+> My Streamlit app connects to Azure Postgres, but when it runs a query on `dev_halyna.fct_trips`, I get this error: `relation "dev_halyna.fct_trips" does not exist`. Is this a problem with the SELECT query, the schema, or the table? Please explain it simply.
 
 ## The response
 
 <!-- TODO: summarise or paste what the LLM returned. -->
 
-TODO
+The LLM explained that the SQL aggregation itself was not the main problem. The error meant that PostgreSQL could not find the table `fct_trips` inside the schema `dev_halyna`.
+
+It suggested checking whether the table exists in the database, whether the app is reading the correct `DB_SCHEMA` from the `.env` file, and whether the Week 10 mart table had been created in my own schema.
 
 ## Reflection
 
 <!-- TODO: what did you change, keep, or discard after reviewing the LLM's answer?
      Be specific: "I kept the cache_data suggestion but changed ttl from 60 to 300
      to match the mart's once-a-day refresh cadence." -->
 
-TODO
+After checking the response, I verified the database table and confirmed that the problem was related to the missing `fct_trips` table in my schema, not to the `COUNT`, `AVG`, or `GROUP BY` logic.
+
+I used the explanation to debug the issue step by step. I kept my Streamlit code close to the Week 11 material, using `run_query`, `@st.cache_data`, `st.metric`, `st.columns`, `st.line_chart`, and `st.sidebar.selectbox`.
+
+I did not paste any real database password, full connection string, or private credentials into the LLM.
 
 ---
 

week11-streamlit/.env.example

@@ -1,4 +1,6 @@
 # PostgreSQL connection string (your Week 9/10 login)
-POSTGRES_URL=postgresql://pipeline_user:your-pg-password@your-pg-host:5432/team1
+# Keep the database name as team1 and the ?sslmode=require suffix; only swap
+# in your own user, password, and host.
+POSTGRES_URL=postgresql://your-pg-user:your-pg-password@your-pg-host:5432/team1?sslmode=require
 # Your dev schema name (e.g. dev_jana)
-DB_SCHEMA=dev_yourname
+DB_SCHEMA=dev_yourname

week11-streamlit/app.py

@@ -18,28 +18,117 @@
 load_dotenv()  # reads .env file if present
 
 POSTGRES_URL = os.environ["POSTGRES_URL"]
-DB_SCHEMA = os.environ.get("DB_SCHEMA", "dev_yourname")
+DB_SCHEMA = os.environ.get("DB_SCHEMA", "dev_halyna")
 
 st.set_page_config(page_title="NYC Taxi Metrics", layout="wide")
 st.title("NYC Taxi Metrics")
 
 
 @st.cache_data(ttl=300)
 def run_query(sql: str) -> pd.DataFrame:
+    """Run a SQL query against the Postgres database and return a DataFrame."""
     engine = sqlalchemy.create_engine(POSTGRES_URL)
     with engine.connect() as conn:
         return pd.read_sql(sql, conn)
 
 
+# Sidebar filter: payment type
+# -----------------------------
+
+st.sidebar.header("Filters")
+
+payment_types = run_query(f"""
+    SELECT DISTINCT payment_type_label
+    FROM {DB_SCHEMA}.fct_trips
+    WHERE payment_type_label IS NOT NULL
+    ORDER BY payment_type_label
+""")["payment_type_label"].tolist()
+
+selected_payment_type = st.sidebar.selectbox(
+    "Payment type",
+    ["All"] + payment_types
+)
+
+if selected_payment_type == "All":
+    WHERE_CLAUSE = ""
+else:
+    WHERE_CLAUSE = f"WHERE payment_type_label = '{selected_payment_type}'"
+
+
+# -----------------------------
+# Headline KPIs
+# -----------------------------
+
 st.subheader("Headline KPIs")
 
-# TODO: query total trip count, average trip_distance, and average
-# fare_per_mile from {DB_SCHEMA}.fct_trips through run_query(), then
-# render three tiles side by side with st.columns(3) and .metric().
-# This is deliberately not the total-trips/avg-fare/total-revenue trio
-# from the chapter: trip_distance and fare_per_mile are different columns,
-# so copying the chapter's SQL verbatim will not answer this.
-raise NotImplementedError(
-    "TODO: implement the headline KPIs panel (total trips, avg trip "
-    "distance, avg fare per mile) from fct_trips."
+kpis = run_query(f"""
+    SELECT
+        COUNT(*) AS total_trips,
+        AVG(trip_distance) AS avg_trip_distance,
+        AVG(fare_per_mile) AS avg_fare_per_mile
+    FROM {DB_SCHEMA}.fct_trips
+    {WHERE_CLAUSE}
+""").iloc[0]
+
+col1, col2, col3 = st.columns(3)
+
+col1.metric(
+    "Total trips",
+    f"{int(kpis['total_trips']):,}"
+)
+
+col2.metric(
+    "Average trip distance",
+    f"{kpis['avg_trip_distance']:.2f} miles"
+)
+
+col3.metric(
+    "Average fare per mile",
+    f"${kpis['avg_fare_per_mile']:.2f}"
+)
+
+
+# -----------------------------
+# Hour-of-day trend
+# -----------------------------
+
+st.subheader("Trips by pickup hour")
+
+hourly = run_query(f"""
+    SELECT
+        EXTRACT(HOUR FROM pickup_datetime)::int AS pickup_hour,
+        COUNT(*) AS trip_count
+    FROM {DB_SCHEMA}.fct_trips
+    {WHERE_CLAUSE}
+    GROUP BY 1
+    ORDER BY 1
+""")
+
+st.line_chart(hourly.set_index("pickup_hour"))
+
+
+# -----------------------------
+# Data freshness
+# -----------------------------
+
+st.subheader("Data freshness")
+
+fresh = run_query(f"""
+    SELECT
+        COUNT(*) AS row_count,
+        MAX(pickup_datetime) AS last_pickup
+    FROM {DB_SCHEMA}.fct_trips
+    {WHERE_CLAUSE}
+""").iloc[0]
+
+col1, col2 = st.columns(2)
+
+col1.metric(
+    "Row count",
+    f"{int(fresh['row_count']):,}"
+)
+
+col2.metric(
+    "Last pickup",
+    str(fresh["last_pickup"])[:16]
 )

week11-streamlit/metric_definitions.md

@@ -0,0 +1,72 @@
+# Metric definitions
+
+Five fields per metric: Name, Description, Calculation, Data source, Refresh frequency. One block per panel. Copy this file to `metric_definitions.md` inside your `week11-streamlit/` folder and fill it in.
+
+## Metabase panels
+
+<!-- One block per Question: trip count by payment type, average fare per
+     mile by dropoff borough, average trip duration by hour of day. -->
+
+### Panel 1: Trip Count by Payment Type
+
+- **Name**: `trip_count_by_payment_type`
+- **Description**: Number of taxi trips grouped by payment type. This shows which payment methods are used most often in the dataset.
+- **Calculation**: `COUNT(*)` grouped by `payment_type_label`
+- **Data source**: `dev_halyna.fct_trips`
+- **Refresh frequency**: Updated when the `fct_trips` mart is rebuilt. Metabase reads the current database table when the dashboard is opened.
+
+### Panel 2: Average Fare per Mile by Dropoff Borough
+
+- **Name**: `avg_fare_per_mile_by_dropoff_borough`
+- **Description**: Average fare charged per mile, grouped by the borough where the trip ended. Unit: US dollars per mile.
+- **Calculation**: `AVG(fare_per_mile)` grouped by `dropoff_borough`
+- **Data source**: `dev_halyna.fct_trips`
+- **Refresh frequency**: Updated when the `fct_trips` mart is rebuilt. Metabase reads the current database table when the dashboard is opened.
+
+### Panel 3: Average Trip Duration by Hour
+
+- **Name**: `avg_trip_duration_by_pickup_hour`
+- **Description**: Average trip duration in minutes, grouped by pickup hour of day. This shows how trip duration changes during the day.
+- **Calculation**: `AVG(EXTRACT(EPOCH FROM (dropoff_datetime - pickup_datetime)) / 60)` grouped by `EXTRACT(HOUR FROM pickup_datetime)`
+- **Data source**: `dev_halyna.fct_trips`
+- **Refresh frequency**: Updated when the `fct_trips` mart is rebuilt. Metabase reads the current database table when the dashboard is opened.
+
+## Streamlit panels
+
+<!-- Headline KPIs panel: total trips, average trip distance, average
+     fare per mile. -->
+
+### Panel 1: Headline KPIs
+
+- **Name**: `headline_kpis`
+- **Description**: Three headline metrics for the selected payment type: total trips, average trip distance, and average fare per mile.
+- **Calculation**:
+  - `total_trips`: `COUNT(*)`
+  - `avg_trip_distance`: `AVG(trip_distance)`
+  - `avg_fare_per_mile`: `AVG(fare_per_mile)`
+  - If a payment type is selected, the query adds `WHERE payment_type_label = selected_payment_type`.
+- **Data source**: `dev_halyna.fct_trips`
+- **Refresh frequency**: Streamlit caches the query result for 300 seconds using `@st.cache_data(ttl=300)`. The underlying data updates when the `fct_trips` mart is rebuilt.
+
+### Panel 2: Trips by Pickup Hour
+
+- **Name**: `trip_count_by_pickup_hour`
+- **Description**: Number of taxi trips grouped by pickup hour of day. This shows the daily demand pattern across the 24-hour cycle.
+- **Calculation**: `COUNT(*)` grouped by `EXTRACT(HOUR FROM pickup_datetime)`. If a payment type is selected, the query adds `WHERE payment_type_label = selected_payment_type`.
+- **Data source**: `dev_halyna.fct_trips`
+- **Refresh frequency**: Streamlit caches the query result for 300 seconds using `@st.cache_data(ttl=300)`. The underlying data updates when the `fct_trips` mart is rebuilt.
+
+### Panel 3: Data Freshness
+
+- **Name**: `data_freshness`
+- **Description**: Data-quality panel showing the number of rows available and the latest pickup timestamp in the selected dataset.
+- **Calculation**:
+  - `row_count`: `COUNT(*)`
+  - `last_pickup`: `MAX(pickup_datetime)`
+  - If a payment type is selected, the query adds `WHERE payment_type_label = selected_payment_type`.
+- **Data source**: `dev_halyna.fct_trips`
+- **Refresh frequency**: Streamlit caches the query result for 300 seconds using `@st.cache_data(ttl=300)`. The underlying data updates when the `fct_trips` mart is rebuilt.
+
+<!-- Add more ### Panel blocks under either section if you build more (the Required tier adds
+     an hour-of-day trend and a freshness panel to Streamlit; a Metabase date filter is Extra,
+     bonus credit, not required). -->