Python Analysis and Exploratory Analysis

Python Analysis

Python lets you do things that are awkward in SQL: quick statistics, charts, and more flexible reshaping. Set up a notebook in the notebooks folder with clear sections, each starting with a text cell: 1. Load data, 2. Check, 3. Analysis by question, 4. Findings.

import pandas as pd
import matplotlib.pyplot as plt
df = pd.read_csv("../data/clean/rides_clean.csv", parse_dates=["start_time"])
df.groupby("rider_type")["ride_minutes"].describe()
df["weekday"] = df["start_time"].dt.day_name()
pd.pivot_table(df, values="ride_id", index="weekday", columns="rider_type", aggfunc="count")

Cross-check: reproduce one or two of your SQL results in pandas. If the numbers differ, find out why before going further.

Exploratory Analysis

Now follow the EDA checklist from the Statistics course, using your SMART questions as a guide but staying open to surprises:

• Distributions: histograms and box plots of key numbers (df["ride_minutes"].plot(kind="hist")). Note the shape and outliers.
• Comparisons: bar charts of averages by group.
• Trends: line charts over time. Look for seasonality.
• Relationships: scatter plots and df.corr(numeric_only=True).

For every chart, write one sentence underneath answering "So what?" For example: "Casual riders take rides 2.3 times longer than members, mostly on weekends." These sentences become your insights next week.

Keep a running list of new questions and limitations (for example "no data on rider age"). Mentioning limitations honestly makes your work more credible.