Skip to main content
added 8 characters in body
Source Link
Buschi Sergio
  • 4.2k
  • 1
  • 22
  • 26

From (Italian, very nice book):"Lezioni di Geometria Analitica e Proiettiva" by Beltrametti, CArletti, Gallarati, Bragadin (pag. 21):

Let $K$ a field, $V:= K^{n+1}$ and let $e_1,\ldots, e_{n+1}$ a base (canonical or not) of $V$. Let $W\subset K^{n+1}$ a $K$-vectorial subspace with dimentiondimension $r+1$, and let $v_1,\ldots, v_{r+1}$ a base of $W$, with

$v_m= a^1_m\cdot e_1 + \ldots a^{n+1}_m\cdot e_n$$v_m= a^1_m\cdot e_1 + \ldots a^{n+1}_m\cdot e_{n+1}$ for $1\leq m\leq r+1$

Let $M$ the matrix with ($n+1$) row's:

$x_1, a^1_1\ldots, a^1_{r+1} $

$x_2, a^2_1\ldots, a^2_{r+1} $

$\ldots, \ldots, \ldots$

$\ldots, \ldots, \ldots$

$x_{n+1}, a^{n+1}_1, \ldots a^{n+1}$

(the last element is $a^{n+1}_{r+1}$)

The book assert (mentioning Kronecker theorem) that

the $r+2$-minor's of $M$ (these are $\binom{n+1}{r+2}$)

considered as linear forms (grade 1 homogeneous polynomial) on variables $x_1,\ldots, x_n$$x_1,\ldots, x_{n+1}$

are linearly dependent, and there are $n-r$ (and no more) linearly independent $r+2$-minors.

Is this true?

How to prove this?

From (Italian, very nice book):"Lezioni di Geometria Analitica e Proiettiva" by Beltrametti, CArletti, Gallarati, Bragadin (pag. 21):

Let $K$ a field, $V:= K^{n+1}$ and let $e_1,\ldots, e_{n+1}$ a base (canonical or not) of $V$. Let $W\subset K^{n+1}$ a $K$-vectorial subspace with dimention $r+1$, and let $v_1,\ldots, v_{r+1}$ a base of $W$, with

$v_m= a^1_m\cdot e_1 + \ldots a^{n+1}_m\cdot e_n$ for $1\leq m\leq r+1$

Let $M$ the matrix with ($n+1$) row's:

$x_1, a^1_1\ldots, a^1_{r+1} $

$x_2, a^2_1\ldots, a^2_{r+1} $

$\ldots, \ldots, \ldots$

$\ldots, \ldots, \ldots$

$x_{n+1}, a^{n+1}_1, \ldots a^{n+1}$

(the last element is $a^{n+1}_{r+1}$)

The book assert (mentioning Kronecker theorem) that

the $r+2$-minor's of $M$ (these are $\binom{n+1}{r+2}$)

considered as linear forms (grade 1 homogeneous polynomial) on variables $x_1,\ldots, x_n$

are linearly dependent, and there are $n-r$ (and no more) linearly independent $r+2$-minors.

Is this true?

How to prove this?

From (Italian, very nice book):"Lezioni di Geometria Analitica e Proiettiva" by Beltrametti, CArletti, Gallarati, Bragadin (pag. 21):

Let $K$ a field, $V:= K^{n+1}$ and let $e_1,\ldots, e_{n+1}$ a base (canonical or not) of $V$. Let $W\subset K^{n+1}$ a $K$-vectorial subspace with dimension $r+1$, and let $v_1,\ldots, v_{r+1}$ a base of $W$, with

$v_m= a^1_m\cdot e_1 + \ldots a^{n+1}_m\cdot e_{n+1}$ for $1\leq m\leq r+1$

Let $M$ the matrix with ($n+1$) row's:

$x_1, a^1_1\ldots, a^1_{r+1} $

$x_2, a^2_1\ldots, a^2_{r+1} $

$\ldots, \ldots, \ldots$

$\ldots, \ldots, \ldots$

$x_{n+1}, a^{n+1}_1, \ldots a^{n+1}$

(the last element is $a^{n+1}_{r+1}$)

The book assert (mentioning Kronecker theorem) that

the $r+2$-minor's of $M$ (these are $\binom{n+1}{r+2}$)

considered as linear forms (grade 1 homogeneous polynomial) on variables $x_1,\ldots, x_{n+1}$

are linearly dependent, and there are $n-r$ (and no more) linearly independent $r+2$-minors.

Is this true?

How to prove this?

Source Link
Buschi Sergio
  • 4.2k
  • 1
  • 22
  • 26

Linear (in)dependence of minors of a matrix

From (Italian, very nice book):"Lezioni di Geometria Analitica e Proiettiva" by Beltrametti, CArletti, Gallarati, Bragadin (pag. 21):

Let $K$ a field, $V:= K^{n+1}$ and let $e_1,\ldots, e_{n+1}$ a base (canonical or not) of $V$. Let $W\subset K^{n+1}$ a $K$-vectorial subspace with dimention $r+1$, and let $v_1,\ldots, v_{r+1}$ a base of $W$, with

$v_m= a^1_m\cdot e_1 + \ldots a^{n+1}_m\cdot e_n$ for $1\leq m\leq r+1$

Let $M$ the matrix with ($n+1$) row's:

$x_1, a^1_1\ldots, a^1_{r+1} $

$x_2, a^2_1\ldots, a^2_{r+1} $

$\ldots, \ldots, \ldots$

$\ldots, \ldots, \ldots$

$x_{n+1}, a^{n+1}_1, \ldots a^{n+1}$

(the last element is $a^{n+1}_{r+1}$)

The book assert (mentioning Kronecker theorem) that

the $r+2$-minor's of $M$ (these are $\binom{n+1}{r+2}$)

considered as linear forms (grade 1 homogeneous polynomial) on variables $x_1,\ldots, x_n$

are linearly dependent, and there are $n-r$ (and no more) linearly independent $r+2$-minors.

Is this true?

How to prove this?