3b1b/manim · error · IndexError

Index {index} out of bound for matrix with {len(self.columns

Error message

Index {index} out of bound for matrix with {len(self.columns)} columns

What it means

IndexError from Matrix.get_column (manimlib/mobject/matrix.py:137) with message 'Index {index} out of bound for matrix with {n} columns'. self.columns is the VGroup of column elements built at construction time from the matrix data; get_column validates 0 <= index < len(self.columns) and raises with the offending index and actual column count so you can see how far off you were.

Source

Thrown at manimlib/mobject/matrix.py:137

        else:
            return Tex(str(element), **config)

    def create_brackets(self, rows, v_buff: float, h_buff: float) -> VGroup:
        brackets = Tex("".join((
            R"\left[\begin{array}{c}",
            *len(rows) * [R"\quad \\"],
            R"\end{array}\right]",
        )))
        brackets.set_height(rows.get_height() + v_buff)
        l_bracket = brackets[:len(brackets) // 2]
        r_bracket = brackets[len(brackets) // 2:]
        l_bracket.next_to(rows, LEFT, h_buff)
        r_bracket.next_to(rows, RIGHT, h_buff)
        return VGroup(l_bracket, r_bracket)

    def get_column(self, index: int):
        if not 0 <= index < len(self.columns):
            raise IndexError(f"Index {index} out of bound for matrix with {len(self.columns)} columns")
        return self.columns[index]

    def get_row(self, index: int):
        if not 0 <= index < len(self.rows):
            raise IndexError(f"Index {index} out of bound for matrix with {len(self.rows)} rows")
        return self.rows[index]

    def get_columns(self) -> VGroup:
        return self.columns

    def get_rows(self) -> VGroup:
        return self.rows

    def set_column_colors(self, *colors: ManimColor) -> Self:
        columns = self.get_columns()
        for color, column in zip(colors, columns):
            column.set_color(color)
        return self

View on GitHub (pinned to dee01804d4)

Solutions

1. Use 0-based indices strictly less than len(matrix.get_columns())
2. Derive bounds from the object: for i in range(len(m.get_columns())): m.get_column(i)
3. For the last column use index len(m.get_columns()) - 1, not -1

Example fix

# before
m = Matrix([[1, 2], [3, 4]])
m.get_column(2)   # raises: Index 2 out of bound ... 2 columns
m.get_column(-1)  # also raises (no negative indexing)

# after
m = Matrix([[1, 2], [3, 4]])
last = m.get_column(len(m.get_columns()) - 1)

n_cols = len(matrix.get_columns())
assert 0 <= col_index < n_cols, f"col_index must be in [0, {n_cols})"
column = matrix.get_column(col_index)

def is_valid_column_index(matrix, i: int) -> bool:
    return 0 <= i < len(matrix.get_columns())

When it happens

Trigger: Matrix([[1,2],[3,4]]).get_column(5); negative indices (get_column(-1) fails this check even though Python slicing would accept it); index computed from user data or loop bounds that exceed the number of columns; matrices constructed via DecimalMatrix/Matrix of a shorter dataset than the index assumes.

Common situations: Coloring specific columns (set_column_colors is safe, but manual get_column(i) loops with wrong bounds); code written against a larger matrix later fed a smaller one; off-by-one loop bounds; assuming 1-based indexing (get_column(1) for the first column is fine, but get_column(n) for the last is not).

Related errors

• Index {index} out of bound for matrix with {len(self.rows)}
• Input to MobjectMatrix must have at least n_rows * n_cols en
• At least 2 mobjects needed for Union.
• At least 2 mobjects needed for Intersection.
• At least 2 mobjects needed for Exclusion.

AI-assisted analysis of 3b1b/manim@dee01804d4 (2026-08-14). Data as JSON: /api/errors/10c4af724ce55f81. Report an issue: GitHub.