Module 24 · Core

Plot and interpret

Select a plot that answers a clear question and label its units.

Before you start: Use summaries, numeric columns, and sorted dates.

Python loads from this site on your first run.

Each run starts with fresh data. Download any exported files before another run. Your written answers and code drafts are saved on this device; export your work to keep a backup.

M24C01 · Explore · predict · practice

Compare groups with bars

A bar chart compares categories. First calculate one statistic per group, then plot those values. The axis label must say whether the height is a mean, count, or total.

ReadWhat does the height of one bar represent?

PredictWhich campus has the higher mean?

ExplainDoes the chart prove campus causes a score difference?

Read the example

df = students()
result = df.groupby("Campus")["Score"].mean().sort_index()
ax = result.plot.bar(color="#28756a", rot=0, figsize=(6, 3.6))
ax.set(title="Mean score by campus (practice data)", xlabel="Campus", ylabel="Mean score (out of 100)", ylim=(0, 100))
plt.tight_layout()
print(result.round(2))
plt.show()

Verified example output · pandas 2.2.3

Campus
Dubai    76.83
RAK      81.00
Name: Score, dtype: float64

Verified example figure for Compare groups with bars

Watch for this. Use a zero baseline for ordinary bar comparisons so bar length remains meaningful.

Your turn

Plot the number of students in each Program instead, with a count label on the y-axis.

For the code check, keep these result variable names: result. Different valid approaches are welcome.

Your Python code

df = students()
result = df.groupby("Campus")["Score"].mean().sort_index()
ax = result.plot.bar(color="#28756a", rot=0, figsize=(6, 3.6))
ax.set(title="Mean score by campus (practice data)", xlabel="Campus", ylabel="Mean score (out of 100)", ylim=(0, 100))
plt.tight_layout()
print(result.round(2))
plt.show()

Show a hint

Use groupby("Program").size() before plot.bar().

Show one worked solution

df = students()
result = df.groupby("Program").size()
ax = result.plot.bar(color="#28756a", rot=0, figsize=(6, 3.6))
ax.set(title="Students by program (practice data)", xlabel="Program", ylabel="Number of students")
ax.set_ylim(bottom=0)
plt.tight_layout()
print(result)
plt.show()

Both programs have six students, so their bars have the same height.

Compare your explanations

Read. The mean student Score for that campus in the fictional table.
Predict. RAK, with 81.00, compared with Dubai at about 76.83.
Explain. No. These are descriptive summaries from a small fictional dataset, without causal evidence.

One quick check

Why label the y-axis "Mean score"?

What you should be able to explain. Aggregate first, then label exactly what was plotted.

Your own assessment, separate from code checks

M24C02 · Explore · predict · practice

Inspect a distribution

A histogram counts numeric observations within intervals called bins. It shows the distribution of individual values, rather than a summary for each named category. Bin choices can change the visual impression.

ReadWhat is counted along the vertical axis?

PredictHow many scores are in [90, 100)?

ExplainHow is this different from a bar chart of campus means?

Read the example

df = students()
result = pd.cut(df["Score"], bins=[50, 60, 70, 80, 90, 100], right=False).value_counts(sort=False)
ax = df["Score"].plot.hist(bins=[50, 60, 70, 80, 90, 100], color="#28756a", edgecolor="white", figsize=(6, 3.6))
ax.set(title="Distribution of scores (practice data)", xlabel="Score (out of 100)", ylabel="Number of students")
plt.tight_layout()
print(result)
plt.show()

Verified example output · pandas 2.2.3

Score
[50, 60)     1
[60, 70)     2
[70, 80)     3
[80, 90)     4
[90, 100)    2
Name: count, dtype: int64

Verified example figure for Inspect a distribution

Watch for this. Histogram libraries usually include the upper endpoint of the last bin; pd.cut(right=False) does not. These sample scores contain no endpoint value of 100.

Your turn

Use wider bins [50, 75, 100] and compare the count table and visual shape.

For the code check, keep these result variable names: result. Different valid approaches are welcome.

Your Python code

df = students()
result = pd.cut(df["Score"], bins=[50, 60, 70, 80, 90, 100], right=False).value_counts(sort=False)
ax = df["Score"].plot.hist(bins=[50, 60, 70, 80, 90, 100], color="#28756a", edgecolor="white", figsize=(6, 3.6))
ax.set(title="Distribution of scores (practice data)", xlabel="Score (out of 100)", ylabel="Number of students")
plt.tight_layout()
print(result)
plt.show()

Show a hint

Change the bin edges in both pd.cut and plot.hist.

Show one worked solution

df = students()
result = pd.cut(df["Score"], bins=[50, 75, 100], right=False).value_counts(sort=False)
ax = df["Score"].plot.hist(bins=[50, 75, 100], color="#28756a", edgecolor="white", figsize=(6, 3.6))
ax.set(title="Scores with wider bins (practice data)", xlabel="Score (out of 100)", ylabel="Number of students")
plt.tight_layout()
print(result)
plt.show()

The two wider bins contain four scores below 75 and eight scores from 75 up to 100.

Compare your explanations

Read. Students whose scores fall inside each interval.
Predict. 2: 91 and 95.
Explain. A histogram groups numeric values into intervals; campus bars compare named groups.

One quick check

What happens when bin widths change?

What you should be able to explain. Check bin boundaries when interpreting distributions.

Your own assessment, separate from code checks

M24C03 · Explore · predict · practice

Compare two measurements

A scatter plot places one point at each pair of numeric measurements. This can reveal a pattern, an unusual observation, or a gap. Correlation describes association; it does not establish why the pattern occurs.

ReadWhat does one dot represent?

PredictWill the overall pattern slope upward or downward?

ExplainCan this prove that adding one hour causes a fixed score increase?

Read the example

df = students()
result = df[["Study_Hours", "Score"]].corr().loc["Study_Hours", "Score"]
ax = df.plot.scatter(x="Study_Hours", y="Score", color="#28756a", figsize=(6, 3.6))
ax.set(title="Study time and score (practice data)", xlabel="Study hours per week", ylabel="Score (out of 100)")
plt.tight_layout()
print("Pearson correlation:", round(result, 3))
plt.show()

Verified example output · pandas 2.2.3

Pearson correlation: 0.985

Verified example figure for Compare two measurements

Watch for this. Outliers, restricted ranges, and missing data can affect correlation; inspect the actual points.

Your turn

Plot Attendance against Score and calculate their Pearson correlation.

For the code check, keep these result variable names: result. Different valid approaches are welcome.

Your Python code

df = students()
result = df[["Study_Hours", "Score"]].corr().loc["Study_Hours", "Score"]
ax = df.plot.scatter(x="Study_Hours", y="Score", color="#28756a", figsize=(6, 3.6))
ax.set(title="Study time and score (practice data)", xlabel="Study hours per week", ylabel="Score (out of 100)")
plt.tight_layout()
print("Pearson correlation:", round(result, 3))
plt.show()

Show a hint

Replace Study_Hours with Attendance in both the correlation and plot.

Show one worked solution

df = students()
result = df[["Attendance", "Score"]].corr().loc["Attendance", "Score"]
ax = df.plot.scatter(x="Attendance", y="Score", color="#28756a", figsize=(6, 3.6))
ax.set(title="Attendance and score (practice data)", xlabel="Attendance (%)", ylabel="Score (out of 100)")
plt.tight_layout()
print("Pearson correlation:", round(result, 3))
plt.show()

The revised plot compares attendance percentages with scores. Its association still does not establish causation.

Compare your explanations

Read. One student with a Study_Hours value and a Score value.
Predict. Upward in this fictional table: higher study hours generally accompany higher scores.
Explain. No. The table is fictional and descriptive, and the plot does not control other factors.

One quick check

What does a positive correlation indicate?

What you should be able to explain. Use scatter plots to inspect relationships without claiming causes.

Your own assessment, separate from code checks

M24C04 · Explore · predict · practice

Show change over time

A line chart uses an ordered horizontal axis to show change. Reindexing to a daily grid keeps the missing date visible as a gap. A line across an unobserved day can otherwise suggest continuity that was not measured.

ReadWhy convert Date before plotting?

PredictWhere will the gap appear?

ExplainDoes the gap mean no visitors came?

Read the example

df = attendance()
df["Date"] = pd.to_datetime(df["Date"], format="%Y-%m-%d")
result = df.set_index("Date")["Visitors"].sort_index().asfreq("D")
fig, ax = plt.subplots(figsize=(6.5, 3.6))
ax.plot(result.index, result, marker="o", color="#28756a")
ax.set(title="Daily visitors (practice data)", xlabel="Date", ylabel="Recorded visitors")
ax.set_ylim(bottom=0)
fig.autofmt_xdate()
plt.tight_layout()
print(result)
plt.show()

Verified example output · pandas 2.2.3

Date
2026-10-01    20.0
2026-10-02    24.0
2026-10-03    18.0
2026-10-04     NaN
2026-10-05    30.0
2026-10-06    28.0
2026-10-07    32.0
Freq: D, Name: Visitors, dtype: float64

Verified example figure for Show change over time

Watch for this. A smooth line does not create new observations; report how any smoothing was calculated.

Your turn

Store the daily counts in daily and the three-day rolling mean in result, requiring three known daily values. Plot daily and result together with labels for both lines.

For the code check, keep these result variable names: result. Different valid approaches are welcome.

Your Python code

df = attendance()
df["Date"] = pd.to_datetime(df["Date"], format="%Y-%m-%d")
result = df.set_index("Date")["Visitors"].sort_index().asfreq("D")
fig, ax = plt.subplots(figsize=(6.5, 3.6))
ax.plot(result.index, result, marker="o", color="#28756a")
ax.set(title="Daily visitors (practice data)", xlabel="Date", ylabel="Recorded visitors")
ax.set_ylim(bottom=0)
fig.autofmt_xdate()
plt.tight_layout()
print(result)
plt.show()

Show a hint

Calculate result from daily.rolling(3, min_periods=3).mean(), then plot both Series on the same axes.

Show one worked solution

df = attendance()
df["Date"] = pd.to_datetime(df["Date"], format="%Y-%m-%d")
daily = df.set_index("Date")["Visitors"].sort_index().asfreq("D")
result = daily.rolling(3, min_periods=3).mean()
fig, ax = plt.subplots(figsize=(6.5, 3.6))
ax.plot(daily.index, daily, marker="o", label="Daily count", color="#28756a")
ax.plot(result.index, result, marker="s", label="3-day mean", color="#9c6b28")
ax.set(title="Daily visitors and rolling mean (practice data)", xlabel="Date", ylabel="Visitors")
ax.set_ylim(bottom=0)
ax.legend()
fig.autofmt_xdate()
plt.tight_layout()
print(result.round(2))
plt.show()

The rolling mean is available only on October 3 and October 7 because other windows lack three known values.

Compare your explanations

Read. It gives the horizontal axis chronological meaning and supports a regular daily index.
Predict. October 4, whose visitor count is missing.
Explain. No. The count is unknown for that day.

One quick check

Why leave a visible gap for the missing day?

What you should be able to explain. Time order, gaps, and transformations belong in the explanation.

Your own assessment, separate from code checks

Previous: Work efficiently and reproducibly Next: Debugging clinic

Further detail: official documentation for this topic.