Summary  
This chapter covers advanced integer-location based indexing using `iloc[]`, allowing selection of specific rows and columns through individual indices, slices, or lists of indices.

General domain of usage  
Data analysis with tabular datasets.

We will learn some new features that `iloc[]` provides. The coolest one is that we can specify indices of both rows and columns. This attribute is similar to `.loc[]`, but the last index of the slicing is exclusive.

Look at the example and the relevant output:

- `data.iloc[1, 2]` - extracts the item located in the dataset's second row and third column. The first index corresponds to the row index, and the second to the column index. Indeed, you can skip one of them;
- `data.iloc[:, 3]` - extracts all values from the rows of the **fourth** column `'IMDb-Rating'`;
- `data.iloc[3, :]` or `data.iloc[3]` - extracts the `4th` row and all relevant columns;
- `data.iloc[:2, 1:4]` - extracts the first two rows and column with the indices `1`, `2`, `3`;
- `data.iloc[[2,4],[1,3]]` - extracts the **rows** with indices `2`,`4` and columns with the indices `1`, `3`.

import unittest
import pandas as pd
import io
import sys


def _dynamic_test(test_case, condition, success_msg, failure_msg):
    if condition:
        test_case._testMethodName = success_msg
        test_case.assertTrue(True, success_msg)
    else:
        test_case._testMethodName = failure_msg
        test_case.fail(failure_msg)


class TestCSVRead(unittest.TestCase):
    def test_csv_read(self):
        """
        1. Read the CSV file into 'data' DataFrame.
        """
        import user_code
        condition = hasattr(user_code, "data") and isinstance(user_code.data, pd.DataFrame)
        _dynamic_test(
            self,
            condition,
            "The CSV file is read correctly into the 'data' DataFrame.",
            "The CSV file is not read correctly or not assigned to 'data'."
        )


class TestDataExtraction(unittest.TestCase):
    def test_data_extracted_slice(self):
        """
        2. Extract first 50 rows and columns with indices 1 and 4 using iloc.
        """
        import user_code
        df = user_code.data
        ref = df.iloc[:50, [1, 4]]
        condition = (
            hasattr(user_code, "data_extracted")
            and isinstance(user_code.data_extracted, pd.DataFrame)
            and user_code.data_extracted.equals(ref)
        )
        _dynamic_test(
            self,
            condition,
            "The variable 'data_extracted' contains the correct rows and columns slice.",
            "The 'data_extracted' slice is incorrect. Check row and column indices."
        )


class TestOutput(unittest.TestCase):
    def test_output_print(self):
        """
        3. Print the first few rows of data_extracted using .head().
        """
        import user_code
        captured_output = io.StringIO()
        sys.stdout = captured_output
        print(user_code.data_extracted.head())
        sys.stdout = sys.__stdout__

        output_text = captured_output.getvalue().strip()
        condition = len(output_text) > 0
        _dynamic_test(
            self,
            condition,
            "The output displays the first few rows of 'data_extracted' using .head().",
            "The 'data_extracted' variable is not printed correctly using .head()."
        )


if __name__ == "__main__":
    unittest.main()


Advanced .iloc[]

Ratkaisu