Q.Solved Case Study based on Open Datasets: UCI dataset is a collection of open datasets, available to the public for experimentation and research purposes. 'auto-mpg' is one such open dataset. It contains data related to fuel consumption by automobiles in a city, measured in miles per gallon (mpg). The data has 398 rows (items/instances) and nine columns (attributes): mpg, cylinders, displacement, horsepower, weight, acceleration, model year, origin, car name. Three attributes — cylinders, model year and origin — have categorical values; car name is a string with a unique value for every row; the remaining five attributes have numeric values. The data is downloaded from the UCI data repository at http://archive.ics.uci.edu/ml/machine-learning-databases/auto-mpg/. Load auto-mpg.data into a DataFrame autodf.

This is a plain theory question — it asks you to explain the process of loading a specific dataset into a pandas DataFrame, which is a core data preprocessing step.

Loading the auto-mpg Dataset into a DataFrame

The UCI auto-mpg dataset is a space-separated file with no header row. You need to load it into a pandas DataFrame while supplying column names and handling missing values.

import pandas as pd

# Define column names as given in the problem
columns = ['mpg', 'cylinders', 'displacement', 'horsepower', 'weight',
           'acceleration', 'model_year', 'origin', 'car_name']

# Load the data — space-separated, no header, handle missing values
autodf = pd.read_csv('auto-mpg.data', sep='\s+', header=None,
                     names=columns, na_values='?')

Key lines explained:

• sep='\s+' — The data uses variable whitespace (spaces or tabs) as delimiters. The regex \s+ matches one or more whitespace characters, correctly splitting each row into columns.
• header=None — The file has no column names in the first row, so you must supply them manually.
• names=columns — Assigns the nine attribute names from the problem statement to the DataFrame columns.
• na_values='?' — The horsepower column contains some ? entries representing missing data. This parameter converts them to NaN so pandas can handle them properly during preprocessing.

After loading, autodf is a DataFrame with 398 rows and 9 columns. The three categorical attributes (cylinders, model_year, origin) are stored as integers, car_name as a string, and the remaining five as floats (with horsepower having some NaN values).

Understanding the file format

The UCI auto-mpg dataset is a classic regression dataset for predicting fuel efficiency. The raw auto-mpg.data file has no header row, and its fields are separated by runs of whitespace (the columns only look fixed-width because the numbers are right-padded for readability — the actual separator is one-or-more spaces, not a fixed character position). Each car's name is the last field and is wrapped in double quotes (e.g. "chevrolet chevelle malibu") specifically because it can itself contain spaces — without the quotes, a plain whitespace split would shred a multi-word car name into several extra columns.

Missing values in the file are marked with a literal ? (this happens in the horsepower column).

The question asks us to load this into a DataFrame called autodf. This is a pure data-loading task, not an analysis one — the challenge is telling pandas how to split the whitespace-separated fields, supply the missing header, and treat ? as missing.

The right tool: pd.read_csv() with a whitespace separator

import pandas as pd

# Column names as given in the problem
columns = ['mpg', 'cylinders', 'displacement', 'horsepower', 'weight',
           'acceleration', 'model_year', 'origin', 'car_name']

# Load the whitespace-separated file
autodf = pd.read_csv(
    'auto-mpg.data',
    sep=r'\s+',
    header=None,
    names=columns,
    na_values='?'
)

What the code does, step by step

1. sep=r'\s+' — The regex \s+ matches one or more whitespace characters (spaces or tabs) between fields, so it correctly splits each line regardless of how many spaces separate two values. Because pandas' parser respects the double quotes around car_name, the spaces inside a quoted car name are not treated as separators — the whole quoted phrase stays one field.
2. header=None — The file has no column-name row, so pandas must not treat the first data row as a header.
3. names=columns — Supplies the nine attribute names from the problem statement, in file order: mpg, cylinders, displacement, horsepower, weight, acceleration, model year, origin, car name.
4. na_values='?' — Converts every ? entry to NaN so the column keeps a numeric dtype.

Verifying the load

print(autodf.shape)        # Should be (398, 9)
print(autodf.dtypes)       # horsepower should be float64, not object
print(autodf.head())       # Preview first 5 rows

Output (first 5 rows):

    mpg  cylinders  displacement  horsepower  weight  acceleration  model_year  origin                   car_name
0  18.0          8         307.0       130.0  3504.0          12.0          70       1  chevrolet chevelle malibu
1  15.0          8         350.0       165.0  3693.0          11.5          70       1          buick skylark 320
2  18.0          8         318.0       150.0  3436.0          11.0          70       1         plymouth satellite
3  16.0          8         304.0       150.0  3433.0          12.0          70       1              amc rebel sst
4  17.0          8         302.0       140.0  3449.0          10.5          70       1                ford torino

Handling missing values (optional but good practice)

# Check for missing values
print(autodf.isnull().sum())

# horsepower has 6 missing values in this dataset
autodf['horsepower'] = autodf['horsepower'].fillna(autodf['horsepower'].median())