Q
stringlengths
4
3.96k
A
stringlengths
1
3k
Result
stringclasses
4 values
Theorem 12.2([4, Theorem 4.2]) Let \( M \) be as in (12.53) and let \( s = d - b{a}^{-1}c \) . Assume that \( s\{ 1,4\} \neq \varnothing \) and let \( {s}^{\left( 1,4\right) } \in s\{ 1,4\} \) and \( f = {1}_{l} - {s}^{\left( 1,4\right) }s \) . Then \( M\{ 1,4\} \neq \varnothing \) if and only if \( v = {1}_{m} + {a}^{...
Proof We use the factorization (12.54). Choose \( {A}^{\left( 1,4\right) } = \left\lbrack \begin{matrix} {a}^{-1} & 0 \\ 0 & {s}^{\left( 1,4\right) } \end{matrix}\right\rbrack \), which\ngives \( I - {A}^{\left( 1,4\right) }A = \left\lbrack \begin{array}{ll} 0 & 0 \\ 0 & f \end{array}\right\rbrack \), where \( f = {1}_...
Yes
Theorem 12.3([4, Theorem 4.3]) Let \( M \) be as in (12.53) and let \( s = d - b{a}^{-1}c \) . Assume that \( {s}^{ \dagger } \) exists and set \( e = {1}_{k} - s{s}^{ \dagger } \) and \( f = {1}_{l} - {s}^{ \dagger }s \) . Then \( {M}^{ \dagger } \) exists if and only if both \( u = {1}_{m} + {\left( b{a}^{-1}\right) ...
Proof In view of the factorization (12.54), we have that \( M \) is Moore-Penrose invertible if and only if \( {PA} \) has a \( \{ 1,3\} \) -inverse and \( {AQ} \) has a \( \{ 1,4\} \) -inverse. Now, since \( {s}^{ \dagger } \) exists we can consider \( {s}^{\left( 1,3\right) } = {d}^{ \dagger } \) in Theorem 12.1 and ...
Yes
Theorem 12.5( \( {}^{\left( 4,\text{ Theorem 4.5 }\right) } \) Let \( e, f \), and \( c \) be as in (12.64) and \( T \) be as in (12.65). Assume that \( c\{ 1,4\} \neq \varnothing \) and let \( {c}^{\left( 1,4\right) } \in c\{ 1,4\} \) . Then \( T\{ 1,4\} \neq \varnothing \) if and only if \( v = {1}_{l} + {d}^{\left( ...
\[ {T}^{\left( 1,4\right) } = \left\lbrack \begin{matrix} {a}^{\left( 1,4\right) } - {hf}{\left( {d}^{\left( 1,2,4\right) }b\right) }^{ * }\mu & {hf}{\left( {d}^{\left( 1,2,4\right) }b\right) }^{ * }\rho + {c}^{\left( 1,4\right) }e \\ - \mu & \rho \end{matrix}\right\rbrack \left\lbrack \begin{matrix} {1}_{m} & 0 \\ - b...
Yes
Proposition 12.6( \( {}^{\left( \left\lbrack 4,\text{Proposition 4.6}\right\rbrack \right) } \) Let \( e, f \) and \( c \) be as in (12.67). If any of the following conditions hold, then \( {c}^{ \dagger } \) exists.\n\n(1) \( w = c{c}^{ * } + d{d}^{ * } \) is invertible.\n\n(2) \( z = {c}^{ * }c + {a}^{ * }a \) is inv...
Proof (1) First, we prove that if \( w \) is invertible then \( c \) is regular. Let \( x = {c}^{ * }{w}^{-1} \) . Since \( {ew} = c{c}^{ * } \) we also have \( e = c{c}^{ * }{w}^{-1} = {cx} \) . Using this, we get \( {cxc} = {ec} = c \) . Hence, \( x \) is a \( \{ 1\} \) -inverse of \( c \) .\n\nNow, choose \( {c}^{\l...
Yes
Theorem 12.7([4, Theorem 4.7]) Let \( T \) be as in (12.60) and let \( e, f \), and \( c \) be as in (12.67). If \( {c}^{ \dagger } \) exists, then \( {T}^{ \dagger } \) exists if and only if both \( u = {1}_{m} + {\left( b{a}^{ \dagger }\right) }^{ * }{egb}{a}^{ \dagger } \) and \( v = {1}_{l} + {d}^{ \dagger }{bhf}{\...
Proof We can apply Theorem 12.4 with \( {a}^{\left( 1,2,3\right) } = {a}^{ \dagger },{d}^{\left( 1,3\right) } = {d}^{ \dagger } \), and \( {c}^{\left( 1,3\right) } = {c}^{ \dagger } \) to obtain that \( {T}^{\left( 1,3\right) } \) exists if and only if \( u = {1}_{m} + {\left( b{a}^{ \dagger }\right) }^{ * }{egb}{a}^{ ...
Yes
Lemma 14. \( {\mathbf{2}}^{\left( \left\lbrack {48},\text{ Lemma }2\right\rbrack \right) } \) Suppose that \( \varphi : X \rightarrow Y \) is a morphism of \( \mathfrak{C} \) with a \( \{ 1,2, i\} \) -inverse \( {\varphi }^{\left( 1,2, i\right) } : Y \rightarrow X \) (with respect to the involution \( * \) ) and that \...
Proof Follows from direct calculation and Jacobson's Lemma.
No
Proposition 14.7( \( {}^{\left( \left\lbrack {48},\text{ Proposition }4\right\rbrack \right) } \) Suppose that \( \varphi : X \rightarrow Y \) is a morphism of \( \mathfrak{C} \) with Moore-Penrose inverse \( {\varphi }^{ \dagger } : Y \rightarrow X \), and that \( \eta : X \rightarrow Y \) is a morphism of \( \mathfra...
Proof The equivalence of conditions (1)-(3) can be obtained from Propositions 14.3 and 14.5.\n\nBy Proposition 14.3, \( {\left( {1}_{X} - \lambda \right) }^{-1}{\left( {1}_{X} + {\varphi }^{ \dagger }\eta \right) }^{-1}{\varphi }^{ \dagger } \in f\{ 1,2,4\} \).\n\nBy Proposition 14.5, \( {\varphi }^{ \dagger }{\left( {...
Yes
Lemma 2.5 If \( \sup \left\{ {{G}_{C} - {\operatorname{pd}}_{R}\left( M\right) \mid M}\right. \) is an \( R \) -module \( \} < \infty \), then for any \( n \geq 0 \) , the following are equivalent:\n\n(1) \( \sup \left\{ {{G}_{C} - {\operatorname{pd}}_{R}\left( M\right) \mid M}\right. \) is an \( R \) -module \( \} \le...
Proof Use [8, Proposition 2.12] and [6, Theorem 9.8].
No
Lemma 3.3 Assume that \( \mathcal{W} \) is a cogenerator for \( \mathcal{X} \) and \( M \in \operatorname{Mod}R \) with \( \mathcal{X} \) - \( {\operatorname{pd}}_{R}\left( M\right) = n < \infty \) . If \( {\operatorname{id}}_{\mathcal{W}}\left( M\right) = 0 \) and \( {\operatorname{id}}_{\mathcal{X}}\left( M\right) < ...
Proof By assumption and \( \left\lbrack {4\text{, Corollary 4.5}}\right\rbrack \), there exists an exact sequence of \( R \) -modules\n\n\[ 0 \rightarrow M \rightarrow N \rightarrow X \rightarrow 0 \]\n\n\( \left( *\right) \)\n\nsuch that \( \mathcal{W} - {\operatorname{pd}}_{R}\left( N\right) \leq n \) and \( X \in \m...
Yes
Corollary 3.4 Let \( M \) be an \( R \) -module with \( {\operatorname{id}}_{R}\left( M\right) < \infty \) and \( {G}_{C} - {\operatorname{pd}}_{R}\left( M\right) < \infty \) . Then \( {\mathcal{P}}_{C} - {\operatorname{pd}}_{R}\left( M\right) = {G}_{C} - {\operatorname{pd}}_{R}\left( M\right) < \infty \) .
Proof It is clear that \( {\operatorname{id}}_{\mathcal{X}}\left( M\right) \leq {\operatorname{id}}_{R}\left( M\right) < \infty \) for any subclass \( \mathcal{X} \) of Mod \( R \) . The finiteness of \( {\operatorname{id}}_{R}\left( M\right) \) implies \( {\operatorname{Ext}}_{R}^{i \geq 1}\left( {C, M}\right) = 0 \) ...
Yes
Theorem 1.3 \( {}^{\left\lbrack 1,\text{ Lemma }2\right\rbrack }\; \) Let \( A \in {\mathbb{C}}^{n \times n} \) with rank \( \left( A\right) = r \) and have representation (1.1). Then the necessary and sufficient condition for the existence of \( {A}^{ \circledast } \) is that \( K \) is nonsingular. Furthermore, when ...
Proof First, observe that \[ {P}_{A} = U\left\lbrack \begin{matrix} {I}_{r} & 0 \\ 0 & 0 \end{matrix}\right\rbrack {U}^{ * } \] Suppose now that \( B \in {\mathbb{C}}^{n \times n} \), partitioned as \[ B = U\left\lbrack \begin{array}{ll} W & X \\ Y & Z \end{array}\right\rbrack {U}^{ * } \] where \( W \in {\mathbb{C}}^{...
Yes
Theorem 1.5 \( {}^{\left\lbrack {32},\text{ Theorem }{2.4}\right\rbrack } \) Let \( A \in {\mathbb{C}}^{n \times n} \) . If \( A = {GH} \) is a full-rank decomposition, then \( {A}^{ \circledast } \) exists if and only if \( {HG} \) is invertible. In this case,
Proof The proof is based on Theorem 1.4 and [3, Theorem 1.7]. Moreover, by \( \left\lbrack 1\right\rbrack ,{A}^{ \circledast } = {A}^{\# }A{A}^{ \dagger } \) . Hence, \( {A}^{ \circledast } = G{\left( HG\right) }^{-1}{\left( {G}^{ * }G\right) }^{-1}{G}^{ * } \) .
Yes
Theorem 2.3 \( {}^{\left\lbrack {29},\text{ Theorem }{2.14}\right\rbrack }\; \) Let \( a \in R \) . The following assertions are equivalent:\n\n(1) \( a \) is core invertible.\n\n(2) There exists \( x \in R \) such that \( \mathrm{{axa}} = a,{}^{ \circ }x = {}^{ \circ }a \) and \( {x}^{ \circ } = {\left( {a}^{ * }\righ...
Proof (1) \( \Rightarrow \) (2). Suppose that \( a \) is core invertible and let \( x = {a}^{ \circledast } \) . By definition, \( {axa} = a,{xR} = {aR} \) and \( {Rx} = R{a}^{ * } \) . By [3, Lemma 2.21], it follows that \( {}^{ \circ }x{ = }^{ \circ }a \) and \( {x}^{ \circ } = {\left( {a}^{ * }\right) }^{ \circ }.\n...
Yes
Theorem 2.6 \( {}^{\left\lbrack {33},\text{ Theorem 2.6 }\right\rbrack } \) Let \( a \in R \) . Then \( a \in {R}^{ \circledast } \) if and only if \( a \in \) \( {R}^{\# } \cap {R}^{\{ 1,3\} } \) . In this case, \( {a}^{ \circledast } = {a}^{\# }a{a}^{\left( 1,3\right) } \) .
Proof Suppose \( a \in {R}^{ \circledast } \) . Then \( {a}^{\# } = {\left( {a}^{ \circledast }\right) }^{2}a \) by Lemma 2.5 and \( a \in {R}^{\{ 1,3\} } \) . Conversely, suppose \( a \in {R}^{\# } \cap {R}^{\{ 1,3\} }. \) Then \( a{a}^{(1,3)}a = a \) and \( (a{a}^{(1,3)}{)}^{ \ast } = a{a}^{(1,3)} \), we have\n\n\[ a...
Yes
Corollary 2. \( {7}^{\left\lbrack {33},\text{ Corollary }{2.7}\right\rbrack } \) Let \( a \in R \) . Then the following conditions are equivalent:\n\n(1) \( a \in {R}^{ \circledast } \) .\n\n(2) \( a \in {R}^{\# } \) and there exists \( x \in R \) such that \( {\left( ax\right) }^{ * } = {ax} \) and \( x{a}^{2} = a \) ...
Proof (1) \( \Rightarrow \) (2). It is clear by Theorem 2.3 and Lemma 2.5.\n\n\( \left( 2\right) \Rightarrow \left( 3\right) . \) It is sufficient to prove that \( x{a}^{2} = a \) implies \( {xa} = {a}^{\# }a \) . It is easy to see that by \( {xa} = x{a}^{2}{a}^{\# } = a{a}^{\# } = {a}^{\# }a \) .\n\n(3) \( \Rightarrow...
Yes
Theorem 2.8 \( {}^{\left\lbrack {33},\text{ Theorem }{3.3}\right\rbrack } \) Let \( a, b \in R \) . If \( {Ra} = R{a}^{2} \), then the following are equivalent:\n\n(1) \( a \in {R}^{ \circledast } \) with core inverse \( b \) .\n\n(2) \( {aba} = a,{\left( ab\right) }^{ * } = {ab} \) and \( a{b}^{2} = b \) .
Proof \( \;\left( 1\right) \Rightarrow \left( 2\right) \) It is trivial by Theorem 2.3.\n\n\( \left( 2\right) \Rightarrow \left( 1\right) \) Suppose that \( {aba} = a, a{b}^{2} = b,{\left( ab\right) }^{ * } = {ab}. \) Then \( a = {aba} = {a}^{2}{b}^{2}a \in {a}^{2}R. \) Thus \( a \in {R}^{\# } \) by the hypothesis \( {...
Yes
Theorem 2.10 \( {}^{\left\lbrack {33},\text{ Theorem }{3.5}\right\rbrack } \) Let \( a, b \in R \) . If \( {aR} = {a}^{2}R \), then the following are equivalent:\n\n(1) \( a \in {R}^{ \circledast } \) with core inverse \( {a}^{ \circledast } = b \) .\n\n(2) \( {bab} = b,{\left( ab\right) }^{ * } = {ab} \) and \( b{a}^{...
Proof \( \left( 1\right) \Rightarrow \left( 2\right) \) . It is trivial by Theorem 2.3.\n\n(2) \( \Rightarrow \) (1). Suppose \( {bab} = b,{\left( ab\right) }^{ * } = {ab} \) and \( b{a}^{2} = a \) . Then \( a = b{a}^{2} \in R{a}^{2} \) , \( a \in {R}^{\# } \) by the hypothesis \( {aR} = {a}^{2}R \) and [14, Propositio...
Yes
Theorem 2.13 \( {}^{\left\lbrack {33},\text{ Theorem }{3.9}\right\rbrack } \) The following conditions are equivalent:\n\n(1) \( R \) is a Dedekind-finite ring.\n\n(2) Let \( a, b \in R \) . Then \( {aba} = a,{\left( ab\right) }^{ * } = {ab}, a{b}^{2} = b \) if and only if \( a \in {R}^{ \circledast } \) with \( {a}^{ ...
Proof \( \;\left( 1\right) \Rightarrow \left( 2\right) \) . Suppose \( a \in {R}^{ \circledast } \) with \( {a}^{ \circledast } = b \) . Then \( {aba} = a,{\left( ab\right) }^{ * } = {ab} \) , \( a{b}^{2} = b \) by Theorem 2.3. Conversely, suppose that \( R \) is a Dedekind-finite ring and \( {aba} = a,{\left( ab\right...
Yes
Theorem 2.14 \( {}^{\left\lbrack {33},\text{ Theorem }{3.10}\right\rbrack } \) Consider the following conditions:\n\n(1) \( R \) is a Dedekind-finite ring.\n\n(2) Let \( a, b \in R \) . If \( {bab} = b,{\left( ab\right) }^{ * } = {ab} \) and \( b{a}^{2} = a \), then \( a \in {R}^{ \oplus } \) with \( {a}^{ \oplus } = b...
Proof \( \;\left( 1\right) \Rightarrow \left( 2\right) \) . Suppose \( R \) is a Dedekind-finite ring. By \( {bab} = b \) and \( b{a}^{2} = a \) , we have\n\n\[ b = {bab} = b\left( {b{a}^{2}}\right) b = {b}^{2}{a}^{2}b \in {b}^{2}R. \]\n\nSince \( b = {bab} \) gives \( a \in b\{ 1\} \), by [4, Theorem 3.1], \( b + 1 - ...
Yes
Proposition 2.15 \( {}^{\left\lbrack {33},\text{ Proposition }{3.11}\right\rbrack }\; \) Let \( a, b \in R \) . If \( {aR} = {a}^{2}R \), then the following conditions are equivalent:\n\n(1) \( a \in {R}^{ \circledast } \) with \( {a}^{ \circledast } = b \) .\n\n(2) \( b{a}^{2} = a,{\left( ab\right) }^{ * } = {ab},{bR}...
Proof \( \;\left( 1\right) \Rightarrow \left( 2\right) \) Suppose \( a \in {R}^{ \circledast } \) with \( {a}^{ \circledast } = b \) . Then \( b{a}^{2} = a,{\left( ab\right) }^{ * } = {ab} \) and \( b = a{b}^{2} \) by Theorem 2.3, thus \( {bR} \subseteq {aR} \) .\n\n(1) \( \Rightarrow \) (3) Suppose \( a \in {R}^{ \cir...
Yes
Proposition 2.16 \( {}^{\left\lbrack {33},\text{ Proposition }{2.11}\right\rbrack } \) Let \( a \in R \) . Then the following conditions are equivalent:\n\n(1) \( a \in {R}^{ \circledast } \) .\n\n(2) \( R = {aR} \oplus {\left( {a}^{ * }\right) }^{ \circ } = {aR} \oplus {a}^{ \circ } \) .\n\n(3) \( R = {aR} + {\left( {...
Proof By Theorem 2.6, we have that \( a \in {R}^{ \circledast } \) if and only if \( a \in {R}^{\# } \cap {R}^{\{ 1,3\} } \) and\n\n\[ {a}^{ \circledast } = {a}^{\# }a{a}^{\left( 1,3\right) }.\]\n\n(2.7)\n\nThus it is easy to see (1)-(9) are equivalent by [3, Theorem 8.11] and [4, Proposition 2.3]. Suppose \( 1 = a{x}_...
Yes
Proposition 2.17 \( {}^{\left\lbrack {33},\text{Proposition 2.12}\right\rbrack }\; \) Let \( a \in R \) . Then the following conditions are equivalent:\n\n(1) \( a \in {R}_{ \circledast } \) .\n\n(2) \( R = {a}^{ * }R \oplus {a}^{ \circ } = {aR} \oplus {a}^{ \circ } \) .\n\n(3) \( R = {a}^{ * }R + {a}^{ \circ } = {aR} ...
In this case,\n\n\[ {a}_{ \circledast } = {x}_{1}^{ * }a{y}_{1} = {x}_{1}^{ * }a{y}_{2}^{2}a = {x}_{2}a{y}_{1} = {x}_{2}a{y}_{2}^{2}a, \]\n\nwhere \( 1 = {a}^{ * }{x}_{1} + {u}_{1} = {x}_{2}a + {u}_{2} = a{y}_{1} + {v}_{1} = {y}_{2}a + {v}_{2} \) for some \( {x}_{1},{x}_{2},{y}_{1},{y}_{2} \in R \) , \( {u}_{2} \in {}^...
Yes
Proposition 2.18 \( {}^{\left\lbrack {19},\text{ Proposition }{2.9}\right\rbrack } \) Let \( a \in R, n \geq 1 \) . We have the following results:\n\n(I) the following conditions are equivalent:\n\n(1) \( a \in R{\left( {a}^{ * }\right) }^{n}a \) .\n\n(2) \( a \in R{a}^{ * }a \cap {a}^{n}R \) .\n\n(3) \( R = {}^{ \circ...
Proof \( \left( I\right) \left( 1\right) \Rightarrow \left( 2\right) . \) It is clear to see that \( a \in R{a}^{ * }a \) follows from \( a \in R{\left( {a}^{ * }\right) }^{n}a, \) and there exists \( r \in R \) such that\n\n\[ a = r{\left( {a}^{ * }\right) }^{n}a = r{\left( {a}^{ * }\right) }^{n - 1}{a}^{ * }a, \]\n\n...
Yes
Theorem 2.19 \( {}^{\left\lbrack {19},\text{ Theorem }{2.10}\right\rbrack }\; \) Let \( a \in R, n \geq 2. \) We have the following results:\n\n(1) \( a \in {R}^{ \circledast } \) if and only if \( a \in R{\left( {a}^{ * }\right) }^{n}a \cap R{a}^{n} \) . In this case, \( {a}^{ \circledast } = {a}^{n - 1}{s}^{ * } \) f...
Proof (1). Since \( a \in {R}^{\# } \) if and only if \( a \in {a}^{2}R \cap R{a}^{2} \) by [4, Theorem 2.2], there exist \( x, y \in R \) such that \( a = {a}^{2}x = y{a}^{2} \) . Further, we have\n\n\[ a = {a}^{2}x = a\left( {{a}^{2}x}\right) x = {a}^{3}{x}^{2} = \cdots = {a}^{n}{x}^{n - 1} \in {a}^{n}R \]\n\nand\n\n...
Yes
Theorem 2.20 \( {}^{\left\lbrack {19},\text{ Theorem 2.11 }\right\rbrack } \) Let \( a \in R, n \geq 2 \) . The following conditions are equivalent:\n\n(1) \( a \in {R}^{ \dagger } \cap {R}^{\# } \) .\n\n(2) \( a \in {R}^{ \circledast } \cap {R}_{ \circledast } \) .\n\n(3) \( a \in a{\left( {a}^{ * }\right) }^{n}R \cap...
Proof The equivalences of seven conditions above and the representations of \( {a}^{ \circledast } \) and \( {a}_{ \circledast } \) can be easily obtained by Proposition 2.18 and Theorem 2.19. We will give the representations of \( {a}^{ \dagger } \) and \( {a}^{\# } \) in the following.\n\nSuppose \( a = s{\left( {a}^...
Yes
Theorem 3.2 \( {}^{\left\lbrack {19},\text{ Theorem }{3.4}\right\rbrack }\; \) Let \( a \in R. \) The following conditions are equivalent:\n\n(1) \( a \in {R}^{ \circledast } \) .\n\n(2) There exists a unique projection \( p \) such that \( {pa} = 0, u = a + p \in {R}^{-1} \) .\n\n(3) There exists an Hermitian element ...
Proof \( \;\left( 1\right) \Rightarrow \left( 2\right) \) . Let \( p = 1 - a{a}^{ \circledast } \) . Then \( p \) is a projection satisfying \( {pa} = 0 \), and the proof of the uniqueness of \( p \) is similar to Theorem 3.1. It is easy to verify\n\n\[ \n\left( {a + 1 - a{a}^{ \circledast }}\right) \left( {{a}^{ \circ...
Yes
Theorem 3.5 \( {}^{\left\lbrack {19},\text{ Theorem }{3.7}\right\rbrack } \) Let \( R \) be a Dedekind-finite ring and \( a \in R \) . The following conditions are equivalent:\n\n(1) \( a \in {R}^{ \circledast } \) .\n\n(2) There exists a unique projection \( p \) such that \( {pa} = 0,{a}^{ * }a + p \in {R}^{-1} \) .\...
Proof Since \( {a}^{ * }a + p \) is Hermitian. Thus \( {a}^{ * }a + p \) is one-sided invertible if and only if it is invertible, hence the conditions (2), (3) and (4) are equivalent. Next, we mainly show the equivalence between the conditions (1) and (2).\n\n(1) \( \Rightarrow \) (2). Assume that \( a \in {R}^{ \circl...
Yes
Theorem 3.6 \( {}^{\left\lbrack {19},\text{ Theorem }{3.8}\right\rbrack } \) Let \( a \in R, n \geq 1 \) . The following conditions are equivalent:\n\n(1) \( a \) is EP.\n\n(2) There exists a unique projection \( p \) such that \( {pa} = {ap} = 0,{a}^{n} + p \in {R}^{-1} \).\n\n(3) There exists an Hermitian element \( ...
Proof (1) \( \Rightarrow \) (2). Suppose \( a \) is EP, so \( a \in {R}^{ \dagger } \cap {R}^{\# } \) and \( {a}^{ \dagger } = {a}^{\# } \) . Let \( p = \) \( 1 - {a}^{\# }a = 1 - {a}^{ \dagger }a \), it is easily seen that \( p \) is a projection satisfying \( {pa} = {ap} = 0 \) . Since\n\n\[ \left( {{a}^{n} + 1 - {a}...
Yes
Theorem 3.7 \( {}^{\left\lbrack 7,\text{ Theorem 4.2 }\right\rbrack } \) Let \( a \in R \) be regular. Then the following conditions are equivalent:\n\n(1) \( a \in {R}^{ \circledast } \) .\n\n(2) \( a + 1 - a{a}^{ - } \) and \( {a}^{ * } + 1 - a{a}^{ - } \) are invertible for some \( {a}^{ - } \in a\{ 1\} \) .\n\n(3) ...
Proof \( \;\left( 1\right) \; \Rightarrow \left( 2\right) .\; \) Since \( a \in {\mathbb{R}}^{ \circledast } \) then \( a \in {\mathbb{R}}^{\# } \cap {\mathbb{R}}^{\left( 1,3\right) } \) by Theorem 2.6. Let \( {a}^{ - } \in a\{ 1,3\} \) . Then \( a + 1 - a{a}^{ - } \) is invertible by [4, Theorem 2.7] and hence \( {a}^...
Yes
Proposition 3.9 \( {}^{\left\lbrack 7,\text{ Proposition }{4.4}\right\rbrack }\; \) Let \( k \geq 1 \) be an integer and suppose that \( a \in R \) is regular. If \( {\left( {a}^{ * }\right) }^{k} + 1 - a{a}^{ - } \in {R}^{-1} \) for any \( {a}^{ - } \in a\{ 1\} \), then \( a \in {R}^{ \oplus } \) .
Proof Let \( u = {\left( {a}^{ * }\right) }^{k} + 1 - a{a}^{ - } \) . As \( u \) is invertible, we have \( a = {u}^{-1}{\left( {a}^{ * }\right) }^{k}a \in R{a}^{ * }a \) . Hence \( a \) is \( \{ 1,3\} \) -invertible by [3, Lemma 8.3].\n\n\[ \text{As}{\left( {\left( {a}^{ * }\right) }^{k} + 1 - a{a}^{\left( 1,3\right) }...
Yes
Let \( R \) be the ring as Remark 3.10. Given \( a = \left\lbrack \begin{matrix} 1 & - 2 \\ 1 & - 2 \end{matrix}\right\rbrack \in R \), then \( {a}^{2} = - a \) and hence \( {a}^{\# } \) exists. So, \( {a}^{\# } \) exists.
Taking \( {a}^{ - } = \left\lbrack \begin{matrix} \frac{2}{3} & \frac{1}{3} \\ 0 & 0 \end{matrix}\right\rbrack \) , then \( {a}^{ * } + 1 - a{a}^{ - } = \frac{1}{3}\left\lbrack \begin{matrix} 4 & 2 \\ - 8 & - 4 \end{matrix}\right\rbrack ,{\left( {a}^{ * }\right) }^{2} + 1 - a{a}^{ - } = \frac{1}{3}\left\lbrack \begin{m...
Yes
Theorem 3.12 \( {}^{\left\lbrack 7,\text{ Theorem }{4.7}\right\rbrack }\; \) Let \( k \geq 1 \) be an integer and suppose \( a \in {R}^{ \oplus } \) . Then the following conditions are equivalent for any \( {a}^{ - } \in a\{ 1\} \) :\n\n(1) \( {\left( {a}^{ * }\right) }^{k} + 1 - a{a}^{ - } \in {R}^{-1} \) .\n\n(2) \( ...
Proof As \( a \in {R}^{ \circledast } \), we get \( a \in {R}^{\# } \) by Theorem 2.6. Hence, \( a + 1 - a{a}^{ \circledast } \in {R}^{-1} \) from \( \left\lbrack {4\text{, Theorem 2.7}}\right\rbrack \) . Note that \( a{a}^{ \circledast } = a{a}^{\left( 1,3\right) } \) and \( {a}^{ * }a{a}^{ \circledast } = {a}^{ * } \...
Yes
Proposition 3.14 \( {}^{\left\lbrack 7,\text{ Proposition 4.9}\right\rbrack } \) Let \( k \geq 1 \) be an integer and suppose \( a \in R \) . Then the following conditions are equivalent:\n\n(1) \( a \in {R}^{ \circledast } \) .\n\n(2) \( a \in {R}^{\left( 1,3\right) } \) and \( {\left( {a}^{ * }\right) }^{k} + 1 - a{a...
Proof \( \left( 1\right) \Rightarrow \left( 2\right) . \) It follows from Theorem 2.6 that \( a \in {\mathbb{R}}^{ \circledast } \) implies \( a \in {\mathbb{R}}^{\# } \cap {\mathbb{R}}^{\left( 1,3\right) }. \) Hence, \( {a}^{k} + 1 - a{a}^{\left( 1,3\right) } \in {R}^{-1} \) from [4, Theorem 2.7].\n\n\[ \text{So,}{\le...
Yes
Proposition 3.18 \( {}^{\left\lbrack 7,\text{ Proposition }{4.13}\right\rbrack }\; \) Let \( a \in {R}^{ \circledast } \) and suppose \( u = {a}^{ * } + 1 - a{a}^{ - } \in \) \( {R}^{-1} \) for some \( {a}^{ - } \in a\{ 1\} \) . Then \( {a}^{ \circledast } = {\left( {u}^{-1}\right) }^{ * }a{\left( {u}^{-1}\right) }^{ *...
Proof \
No
Corollary 3.19 \( {}^{\left\lbrack 7,\text{ Corollary }{4.14}\right\rbrack } \) Let \( R \) be a Dedekind-finite ring. Then the following conditions are equivalent:\n\n(1) \( a \in {R}^{ \circledast } \) .\n\n(2) \( a \in {R}^{\left( 1,3\right) } \) and \( {a}^{ * }a + 1 - a{a}^{\left( 1,3\right) } \) is invertible for...
Proof Let \( u = {a}^{ * } + 1 - a{a}^{\left( 1,3\right) } \) and \( v = {a}^{ * }a + 1 - a{a}^{\left( 1,3\right) } \) . Then \( v = u{u}^{ * } \) . Since \( R \) is a Dedekind-finite ring, we have \( v \in {R}^{-1} \) if and only if \( u \in {R}^{-1} \) . By Corollary 3.16, \( {a}^{ \circledast } = {\left( {u}^{-1}\ri...
Yes
Theorem 3.23 \( {}^{\left\lbrack 7,\text{ Theorem }{5.6}\right\rbrack }\; \) Let \( a \in R \) be regular with an inner inverse \( {a}^{ - } \) . Then the following conditions are equivalent:\n\n(1) \( a \in {R}^{\# } \cap {R}^{ \dagger } \) .\n\n(2) \( a \in {R}^{ \circledast } \cap {R}_{ \circledast } \) .\n\n(3) \( ...
Proof (1) \( \Leftrightarrow \) (2). By Theorem 2.6.\n\n(3) \( \Leftrightarrow \) (5) and (4) \( \Leftrightarrow \) (6). By Jacobson’s lemma.\n\n\( \left( 1\right) \Rightarrow \left( 3\right) . \) From \( \left\lbrack {4,\text{ Theorem }{2.7}}\right\rbrack \) and \( \left\lbrack {3,\text{ Theorem }{8.14}}\right\rbrack ...
Yes
Corollary 3.24 \( {}^{\left\lbrack 7,\text{ Corollary }{5.7}\right\rbrack } \) Let \( a \in {R}^{ \dagger } \) . Then the following conditions are equivalent:\n\n(1) \( a \in {R}^{ \oplus } \) .\n\n(2) \( a \in {R}_{ \oplus } \) .\n\n(3) \( u = a{a}^{ * }a + 1 - a{a}^{ \dagger } \in {R}^{-1} \) .\n\n(4) \( v = a{a}^{ *...
Proof As \( a \in {R}^{ \dagger } \), it follows that \( a \in {R}^{ \circledast } \) if and only if \( a \in {R}^{\# } \) if and only if \( a \in {R}_{ \circledast } \) by Theorem 2.6. So (1)-(6) are equivalent by Theorem 3.23. Moreover, \( {a}^{ \circledast } = {u}^{-1}a{a}^{ * } \) and \( {a}_{ \circledast } = {a}^{...
Yes
Proposition 3.25 \( {}^{\left\lbrack 7,\text{ Proposition }{5.8}\right\rbrack }\; \) Let \( a \in {R}^{ \dagger } \) . Then the following conditions are equivalent:\n\n(1) \( a \in {R}^{ \circledast } \) .\n\n(2) \( a \in {R}^{\# } \) .\n\n(3) \( {a}^{ * } + 1 - a{a}^{ \dagger } \in {R}^{-1} \) .\n\nIn this case, \( {a...
Proof \( \left( 1\right) \Leftrightarrow \left( 2\right) \) . It follows by Theorem 3.23 \( \left( 1\right) \Leftrightarrow \left( 2\right) \) .\n\n(2) \( \Leftrightarrow \) (3). Note that \( {a}^{ * } + 1 - a{a}^{ \dagger } = {\left( a + 1 - a{a}^{ \dagger }\right) }^{ * } \) . It follows from [4, Theorem 2.7] that \(...
Yes
Proposition 3.26 \( {}^{\left\lbrack 7,\text{ Proposition }{5.9}\right\rbrack }\; \) Let \( a \in {R}^{\# } \) . Then the following conditions are equivalent:\n\n(1) \( a \in {R}^{ \circledast } \cap {R}_{ \circledast } \) .\n\n(2) \( a \in {R}^{ \dagger } \) .\n\n(3) \( {a}^{ * } + 1 - a{a}^{\# } \in {R}^{-1} \) .\n\n...
Proof (1) \( \Leftrightarrow \) (2). It follows by Theorem 3.23 (1) \( \Leftrightarrow \) (2).\n\n\( \left( 2\right) \Rightarrow \left( 3\right) . \) Note that \( a \in {R}^{ \dagger } \) implies \( {a}^{ * }a + 1 - {a}^{\# }a \in {R}^{-1} \) by [3, Theorem 8.14]. As \( a \in {R}^{\# } \), then \( a + 1 - a{a}^{ \dagge...
Yes
Lemma 3.32 \( \left\lbrack {{26},\text{ Theorem }{3.2}}\right\rbrack ,\left\lbrack {{25},\text{ Theorem }{1.3}}\right\rbrack \; \) Let \( d \in R \) be regular with \( {d}^{ - } \in d\{ 1\} . \) The following conditions are equivalent:\n\n(1) \( a \) is invertible along \( d \) .\n\n(2) \( u = {da} + 1 - d{d}^{ - } \) ...
In [26, Theorem 3.2], Mary and Patrício proved that \( m \) is invertible along \( d \) if and only if \( d \in {dmdR} \cap {Rdmd} \) . Comparing Theorem 3.29 and Lemma 3.32, we have the following corollary.
No
Proposition 3.35 \( {}^{\left\lbrack {19},\text{ Proposition 4.4 }\right\rbrack } \) Let \( a \in R, n \geq 1 \), we have the following results:\n\n(1) If \( a \in {Ra}{\left( {a}^{ * }\right) }^{n}a \), then \( a \in {a}^{n}{a}^{ * }{a}^{n}R \) .
Proof (1). Suppose \( a \in {Ra}{\left( {a}^{ * }\right) }^{n}a \), there exists \( x \in R \) such that\n\n\[ a = {xa}{\left( {a}^{ * }\right) }^{n}a. \]\n\n(3.17)\n\nTaking involution on (3.17), we get\n\n\[ {a}^{ * } = {a}^{ * }{a}^{n}{a}^{ * }{x}^{ * }. \]\n\n(3.18)\n\nAgain by (3.17), we obtain\n\n\[ {\left\lbrack...
Yes
Theorem 3.38 \( {}^{\left\lbrack {19},\text{ Theorem 4.6 }\right\rbrack } \) Let \( a \in R, n \geq 2 \) . Consider the following conditions:\n\n(1) \( R \) is a Dedekind-finite ring.\n\n(2) \( a \in a{\left( {a}^{ * }\right) }^{n}{aR} \) if and only if \( a \in {Ra}{\left( {a}^{ * }\right) }^{n}a \) .\n\n(3) \( a{a}^{...
Proof \( \;\left( 1\right) \Rightarrow \left( 2\right) . \) Since \( \;R\; \) is a Dedekind-finite ring, it follows that \( \;{\left( {a}^{ * }\right) }^{n}a + 1{-a}^{ - }a \) is right invertible if and only if \( {\left( {a}^{ * }\right) }^{n}a + 1 - {a}^{ - }a \) is left invertible. Hence, \( a \in a{\left( {a}^{ * }...
Yes
Theorem 4. \( {\mathbf{6}}^{\left\lbrack {4.6},\text{ Theorem }{4.8}\right\rbrack } \) Let \( a, b \in {R}_{ \oplus } \) with dual core inverses \( {a}_{ \oplus } \) and \( {b}_{ \oplus }, \) respectively. If \( {ab} = 0 \) and \( a{b}^{ * } = 0 \), then \( a + b \in {R}_{ \oplus } \) . Moreover,
\[ {\left( a + b\right) }_{ \circledast } = {a}_{ \circledast } + {b}_{ \circledast }\left( {1 - a{a}_{ \circledast }}\right) . \]
No
Theorem 4.11 \( {}^{\left\lbrack {39},\text{ Theorem 4.4 }\right\rbrack } \) Let \( a, b \in {R}^{ \oplus } \) . If \( b{a}^{ \oplus }a = a = a{a}^{ \oplus }b \), then \( a - b \in \) \( {R}^{ \circledast } \) and \[ {\left( a - b\right) }^{ \circledast } = a{a}^{ \circledast }{b}^{ \circledast } - {b}^{ \circledast }....
Proof \( \; \) Since \( a = a{a}^{ \circledast }b,\;a{b}^{ \circledast } = a{a}^{ \circledast }b{b}^{ \circledast } \) . According to the proof of Theorem 4.10, we have the following equalities: \[ a{a}^{ \circledast } = a{a}^{ \circledast }b{b}^{ \circledast },\;{b}^{ \circledast }{ba} = a = a{a}^{ \circledast }{b}^{ ...
Yes
Theorem 5. \( {1}^{\left\lbrack {17},\text{ Theorem }{2.6}\right\rbrack } \) Let \( a, p, q \in R \) be such that \( a\{ 1,4\} \neq \varnothing \) . Suppose that there exist \( {p}^{\prime },{q}^{\prime } \in R \) such that \( {p}^{\prime }{pa} = a = {aq}{q}^{\prime } \) . Then the following are equivalent:\n\n(1) \( {...
Proof By Corollary 3.6, \( {paq} \in {R}^{ \circledast } \) if and only if \( \left( {paq}\right) \{ 1,3\} \neq \varnothing \) and \( v = {paq} + 1 - {paq}{\left( paq\right) }^{\left( 1,3\right) } \) is a unit, and in this case \( {\left( paq\right) }^{ \circledast } = {v}^{-1}{paq}{\left( paq\right) }^{\left( 1,3\righ...
Yes
Theorem 5. \( {2}^{\left\lbrack {17},\text{ Theorem }{2.7}\right\rbrack } \) Let \( a, p, q \in R \) be such that \( a\{ 1,3\} \neq \varnothing \) . Suppose that there exist \( {p}^{\prime },{q}^{\prime } \in R \) such that \( {p}^{\prime }{pa} = a = {aq}{q}^{\prime } \) . Then the following are equivalent:\n\n(1) \( {...
Proof In view of [4, Corollary 3.3], \( {\left( paq\right) }^{\# } \) exists if and only if \( u = {aqpa}{a}^{\left( 1,3\right) } + 1 - \) \( a{a}^{\left( 1,3\right) } \) is a unit if and only if \( v = {a}^{\left( 1,3\right) }{aqpa} + 1 - {a}^{\left( 1,3\right) }a \) is a unit when \( {p}^{\prime }{pa} = a = {aq}{q}^{...
Yes
Let \( a, p, q \in R \) be such that \( a\{ 1,4\} \neq \varnothing \) and \( p, q \) be invertible. Then \( {\left( paq\right) }_{ \oplus } \) exists if and only if \( s = {aqp} + 1 - a{a}^{\left( 1,4\right) } \) is a unit and \( \left( {paq}\right) \{ 1,4\} \neq \varnothing \), or, equivalently, \( t = {qpa} + 1 - {a}...
In this case, \[ {\left( paq\right) }_{ \circledast } = {\left( paq\right) }^{\left( 1,4\right) }{paqp}{s}^{-2}{aq} = {\left( paq\right) }^{\left( 1,4\right) }{paqpa}{t}^{-2}q. \]
Yes
Theorem 5. \( {7}^{\left\lbrack {17},\text{ Theorem }{2.12}\right\rbrack } \) Let \( a, p, q \in R \) be such that \( a \in {R}^{ \dagger } \) . Suppose that there exist \( {p}^{\prime },{q}^{\prime } \in R \) such that \( {p}^{\prime }{pa} = a = {aq}{q}^{\prime } \) . Then the following are equivalent:\n\n(1) \( {\lef...
In this case,\n\n\[ {\left( paq\right) }^{ \circledast } = {w}^{-2}{paqpa}{u}^{-1}{\left( pa\right) }^{ * } = {paq}{v}^{-2}{pa}{u}^{-1}{\left( pa\right) }^{ * },\]\n\n\[ {\left( paq\right) }_{ \circledast } = {\left( aq\right) }^{ * }{s}^{-1}{aq}{w}^{-2}{paq} = {\left( aq\right) }^{ * }{s}^{-1}{aqpaq}{v}^{-2}. \]
Yes
Theorem 5. \( {8}^{\left\lbrack {17},\text{ Thoerme }{3.1}\right\rbrack }\; \) Suppose \( a \in R \) is regular with \( {a}^{ - } \in a\{ 1\} \) and \( p, q \in \) \( R \) ,\n\n\[ \nu = {aq}{\left( paq\right) }^{ * }{pa}{a}^{ - } + 1 - a{a}^{ - },\;v = {a}^{ - }{aq}{\left( paq\right) }^{ * }{pa} + 1 - {a}^{ - }a, \]\n\...
Proof From [4, Theorem 3.1], \( {sR} = R \) if and only if \( {aqpaR} = {aR} \) if and only if \( {tR} = R \) . By symmetry it follows that \( {Rs} = R \) is equivalent to \( R \) a \( {qpa} = {Ra} \), or equivalently, \( {Rt} = R \) . So \( s \) is a unit if and only if \( t \) is a unit if and only if \( {Ra} = {Raqp...
Yes
Theorem 6.3. \( {}^{\left( \left\lbrack {13},\text{ Theorem }{2.1}\right\rbrack \right) } \) If \( L = \left\lbrack \begin{matrix} 0 & a \\ {I}_{n - 1} & \beta \end{matrix}\right\rbrack \) with \( \beta = {\left\lbrack {a}_{1},{a}_{2},\cdots ,{a}_{n - 1}\right\rbrack }^{\mathrm{T}} \) is a companion matrix over \( R \)...
Proof Consider the factorization of \( L \) :\n\n\[ \nL = \left\lbrack \begin{matrix} 0 & a \\ {I}_{n - 1} & \beta \end{matrix}\right\rbrack = \left\lbrack \begin{matrix} 0 & 1 \\ {I}_{n - 1} & 0 \end{matrix}\right\rbrack \left\lbrack \begin{matrix} {I}_{n - 1} & 0 \\ 0 & a \end{matrix}\right\rbrack \left\lbrack \begin...
Yes
Theorem 7.1. \( {}^{\left\lbrack {21},\text{ Theorem }{2.1}\right\rbrack }\; \) Let \( \;\mathcal{C} \) be an additive category with an involution \( * \) . Suppose that \( \varphi : X \rightarrow X \) is a morphism of \( \mathcal{C} \) with core inverse \( {\varphi }^{ \circledast } \) and \( \eta : X \rightarrow X \)...
Proof By \( \left\lbrack {3,\text{ Lemma }{14.1}}\right\rbrack ,{\left( {1}_{X} + {\varphi }^{ \circledast }\eta \right) }^{-1}{\varphi }^{ \circledast } \in f\{ 1,2\} .\n\nLet \( {f}_{0} = {\left( {1}_{X} + {\varphi }^{ \circledast }\eta \right) }^{-1}{\varphi }^{ \circledast } = \alpha {\varphi }^{ \circledast } = {\...
Yes
Theorem 7.4. \( {}^{\left\lbrack {21},\text{ Theorem }{3.1}\right\rbrack }\; \) If \( a \in {R}^{ \circledast } \) with core inverse \( {a}^{ \circledast } \) and \( j \in J\left( R\right) \), then\n\n\[ a + j \in {R}^{ \circledast }\text{ if and only if }\varepsilon = \left( {1 - a{a}^{ \circledast }}\right) j{\left( ...
Proof Remark first that, if \( j \in J\left( R\right) \), then \( 1 + {a}^{ \circledast }j \in {R}^{-1} \) and \( {j}^{ * } \in J\left( R\right) \).\n\nSet \( \phi = {\left( a + j\right) }^{ \circledast } \), then \( \phi \in \varepsilon \{ 1\} \) by [3, Lemma 14.1]. This shows that the element \( \varepsilon \in J\lef...
Yes
Theorem 8.2. \( {}^{\left\lbrack {23},\text{ Theorem }{3.5}\right\rbrack }\; \) Let \( A \in {\mathbb{C}}^{n \times n} \) with ind \( \left( A\right) = k \) . Then the core-EP inverse of \( A \) always exists and it is unique. Further, \( {A}^{\circledR } = {A}^{k}{\left( {\left( {A}^{ * }\right) }^{k}{A}^{k + 1}\right...
Proof From the definition, we know that core-EP inverse of \( A \), is an outer inverse of \( A \) with its column and row spaces are identical, and equals \( \mathcal{R}\left( {A}^{k}\right) \) . By [22, Lemma 4.3], we get the core-EP inverse of \( A \) exists if and only if \( \operatorname{rank}\left( {{\left( {A}^{...
Yes
Theorem 8.3. \( {}^{\left\lbrack {11},\text{ Theorem 2.3}\right\rbrack } \) Let \( A \in {\mathbb{C}}^{n \times n} \) with ind \( \left( A\right) = k \) . Then \( {A}^{\text{① }} = \) \( {A}^{D}{A}^{k}{\left( {A}^{k}\right) }^{ \dagger } \) .
Proof Let \( X = {A}^{D}{A}^{k}{\left( {A}^{k}\right) }^{ \dagger } \) . Then it is easy to check that \( X \) is an outer inverse of \( A \) and \( \mathcal{R}\left( X\right) = \mathcal{R}\left( {X}^{ * }\right) = \mathcal{R}\left( {A}^{k}\right) \) .
No
Theorem 8.4. Let \( A \in {\mathbb{C}}^{n \times n} \) be of the form (1.1). Then\n\n\[ \n{A}^{ \oplus } = U\left\lbrack \begin{matrix} {\left( \sum K\right) }^{ \oplus } & 0 \\ 0 & 0 \end{matrix}\right\rbrack {U}^{ * }\n\]
Proof Let \( X = U\left\lbrack \begin{matrix} {\left( \sum K\right) }^{\left( 1\right) } & 0 \\ 0 & 0 \end{matrix}\right\rbrack {U}^{ * } \) . Then it is easy to show that there exits \( k \geq \operatorname{ind}\left( {\sum K}\right) \) such that \( {XAX} = X, X{A}^{k + 1} = {A}^{k} \) and \( {\left( AX\right) }^{ * }...
Yes
Lemma 8.5. \( {}^{\left\lbrack 6\right\rbrack } \) Let \( {\lambda }_{1} \neq {\lambda }_{2}, a \in \mathbb{C} \) . Then \( \left\lbrack \begin{matrix} {\lambda }_{1} & a \\ 0 & {\lambda }_{2} \end{matrix}\right\rbrack \) is unitary similar to \( \left\lbrack \begin{matrix} {\lambda }_{2} & b \\ 0 & {\lambda }_{1} \end...
Proof Let \( k = \frac{a}{\lambda }, Q = \frac{1}{\sqrt{{k}^{ * }k + 1}}\left\lbrack \begin{matrix} k & 1 \\ 1 & - {k}^{ * } \end{matrix}\right\rbrack \) . Then \( Q \) is unitary matrix and \( {Q}^{ * }\left\lbrack \begin{matrix} {\lambda }_{1} & a \\ 0 & {\lambda }_{2} \end{matrix}\right\rbrack Q = \left\lbrack \begi...
Yes
Theorem 8.6 \( {}^{\left\lbrack 6\right\rbrack } \) Let \( A \in {\mathbb{C}}^{n \times n} \) . Then \( A = U\left\lbrack \begin{matrix} D & L \\ 0 & N \end{matrix}\right\rbrack {U}^{ * } \), where \( U \in {\mathbb{C}}^{n \times n} \) is unitary, \( D \in {\mathbb{C}}^{t \times t} \) is nonsingular or zero, \( N \in {...
Proof From Schur’s lemma, there exists a unitary matrix \( U \) such that \( {U}^{ * }{AU} = \n\n\( T \) is a upper triangular matrix. Set \( T = \left\lbrack \begin{matrix} {\lambda }_{1} & {t}_{12} & \cdots & {t}_{1n} \\ & {\lambda }_{2} & \cdots & {t}_{2n} \\ & & \ddots & \vdots \\ & & & {\lambda }_{n} \end{matrix}\...
Yes
Theorem 8.7. \( {}^{\left\lbrack {10},\text{ Theorem 2.7 }\right\rbrack } \) Let \( A \in {\mathbb{C}}^{n \times n} \) with \( \operatorname{ind}\left( A\right) = k \) . Then \( {A}^{\text{① }} = \) \( {\left( A{P}_{{A}^{k}}\right) }^{ \dagger } \) .
Proof Let \( A \in {\mathbb{C}}^{n \times n} \) . Then \( A = U\left\lbrack \begin{matrix} D & L \\ 0 & N \end{matrix}\right\rbrack {U}^{ * } \), and \( {A}^{\left( \oplus \right) } = U\left\lbrack \begin{matrix} {D}^{-1} & 0 \\ 0 & 0 \end{matrix}\right\rbrack {U}^{ * } \) . \n\nSince \( A{P}_{{A}^{k}} = U\left\lbrack ...
No
Theorem 8.8. \( {}^{\left\lbrack {11},\text{ Theorem }{5.2}\right\rbrack } \) Let \( A \in {\mathbb{C}}^{n \times n} \) be of the form (8.69). Then\n\n\[ \n{A}^{ \oplus } = {Q}_{1}{D}^{-1}{\left( {Q}_{1}^{ * }{Q}_{1}\right) }^{-1}{Q}_{1}^{ * }\n\]
Proof From \( P = \left\lbrack \begin{array}{l} {P}_{1} \\ {P}_{2} \end{array}\right\rbrack ,{P}^{-1} = \left\lbrack {{Q}_{1},{Q}_{2}}\right\rbrack \) and \( A = {P}^{-1}\left\lbrack \begin{array}{ll} D & O \\ O & N \end{array}\right\rbrack P \), we have\n\n\[ \n{A}^{D} = {P}^{-1}\left\lbrack \begin{matrix} {D}^{-1} & ...
Yes
Theorem 8.9. Let \( A \in {\mathbb{C}}^{n \times n} \) with \( \operatorname{ind}\left( A\right) = k \) . We perform a sequence of full-rank decompositions:\n\n\[ \nA = {B}_{1}{G}_{1},{G}_{i}{B}_{i} = {B}_{i + 1}{G}_{i + 1},\;i = 1,2,\ldots, k - 1.\n\]\n\nThen\n\[ \n{A}^{ \oplus } = \left\{ \begin{array}{rr} B{\left( {...
Proof Set \( B = {B}_{1}{B}_{2}\cdots {B}_{k} \) and \( G = {G}_{k}\cdots {G}_{2}{G}_{1} \) . Then\n\n\[ \n\begin{matrix} {A}^{\circledR } & = & {A}^{D}{A}^{k}{\left( {A}^{k}\right) }^{ \dagger } = B{\left( {G}_{k}{B}_{k}\right) }^{-k - 1}{GBG}{G}^{ * }{\left( G{G}^{ * }\right) }^{-1}{\left( {B}^{ * }B\right) }^{-1}{B}...
Yes
Lemma 9.3. \( {}^{\left\lbrack {11},\text{ Lemma }{2.1}\right\rbrack }\; \) Let \( a \in R \) . If there exist \( x \in R \) and \( k \in {\mathbb{N}}^{ + } \) such that \( x{a}^{k + 1} = {a}^{k} \) and \( a{x}^{2} = x \), then\n\n(1) \( {ax} = {a}^{m}{x}^{m} \) for arbitrary positive integer \( m \) .\n\n(2) \( x \) a...
Proof (1). Since \( a{x}^{2} = x \), we have \( {ax} = a\left( {a{x}^{2}}\right) = {a}^{2}{x}^{2} = {a}^{2}\left( {a{x}^{2}}\right) x = {a}^{3}{x}^{3} = \) \( \cdots = {a}^{m}{x}^{m} \) for arbitrary positive integer \( m \) .\n\n(2). As \( {ax} = {a}^{k + 1}{x}^{k + 1} = {a}^{k}{x}^{k} \) follows from (1), we have\n\n...
Yes
Theorem 9.4. \( {}^{\left\lbrack {11},\text{ Theorem }{2.2}\right\rbrack } \) Let \( a \in R \) . Then \( a \) has at most one pseudo core inverse in \( R \) .
Proof Suppose \( x \) and \( y \) satisfy condition (1)-(3) of Definition 9.1, with \( m, n \) as pseudo core index respectively. Let \( k = \max \{ m, n\} \) . Then by Lemma 9.3,\n\n\[ \n{x}^{k + 1}{a}^{k} = {x}^{m + 1}{a}^{m} = {a}^{D} = {y}^{n + 1}{a}^{n} = {y}^{k + 1}{a}^{k}, \n\]\n\nand\n\n\[ \n{a}^{k}{x}^{k}{a}^{...
Yes
Theorem 9.5 \( {}^{\left\lbrack {11},\text{ Theorem 2.3}\right\rbrack }\; \) Let \( a \in R \) and \( m \in {\mathbb{N}}^{ + } \) with \( m \geq k. \) Then \( a \in {R}^{\bigodot } \) with \( I\left( a\right) = k \) if and only if \( a \in {R}^{D} \) with \( \operatorname{ind}\left( a\right) = k \) and \( {a}^{m} \in {...
Proof Suppose \( a \in {R}^{ \odot } \) with \( I\left( a\right) = k \) . By Lemma 9.3, we have \( a \in {R}^{D} \) with ind \( \left( a\right) \leq k \) and \( {a}^{m} \in {R}^{\{ 1,3\} } \) for arbitrary \( m \geq k \) . ind \( \left( a\right) < k \) would mean that \( x{a}^{k} = \) \( {a}^{k - 1} \), which is contra...
Yes
Theorem 9.7. \( {}^{\left\lbrack {11},\text{ Theorem }{2.5}\right\rbrack }\; \) Let \( a \in R. \) Then \( a \in {R}^{\circledR } \) if and only if \( {a}^{m} \in {R}^{\circledR } \) for some positive integer \( m \) . In this case, \( {\left( {a}^{m}\right) }^{ \circledast } = {\left( {a}^{\circledR }\right) }^{m} \) ...
Proof Suppose \( {a}^{ \odot } = x \) with \( I\left( a\right) = m \) . Setting \( y = {x}^{m} \), by the definition of pseudo core inverse and by Lemma 9.3, we can check that\n\n\[ y{\left( {a}^{m}\right) }^{2} = {x}^{m}{\left( {a}^{m}\right) }^{2} = \left( {{x}^{m}{a}^{m}}\right) {a}^{m} = \left( {{x}^{m + 1}{a}^{m +...
Yes
Theorem 9.9. \( {}^{\left\lbrack {11},\text{ Theorem }{2.7}\right\rbrack } \) Let \( a \in R \) . If \( a \in {R}^{\circledR } \), then \( {a}^{\circledR } \in {R}^{\circledR } \) . In fact \( {a}^{\circledR } \) is core invertible whenever it exists, and \( {\left( {a}^{\circledR }\right) }^{\circledR } = {\left( {a}^...
Proof To prove this, one has merely to verify that if \( x \) satisfies condition (1)-(3) of Definition 9.1, then \( y = {a}^{2}x \) satisfies\n\n\[ \ny{x}^{2} = x, x{y}^{2} = y,{\left( xy\right) }^{ * } = {xy}.\n\]\n\nHere we omit the details.
No
Proposition 9.10. \( {}^{\left\lbrack {11},\text{ Proposition }{2.8}\right\rbrack } \) Let \( a \in {R}^{\circledR } \) . Then \( {\left( {\left( {a}^{\circledR }\right) }^{\circledR }\right) }^{\circledR } = {a}^{\circledR } \).
Proof Suppose \( a \in {R}^{ \odot } \) with \( I\left( a\right) = m \) . By Theorem 9.9, we have\n\n\[ \n{\left( {\left( {a}^{ \oplus }\right) }^{ \oplus }\right) }^{ \oplus } = {\left( {a}^{ \oplus }\right) }^{2}{\left( {a}^{ \oplus }\right) }^{ \oplus } = {\left( {a}^{ \oplus }\right) }^{2}{a}^{2}{a}^{ \oplus }.\n\]...
Yes
Theorem 9.11. \( {}^{\left\lbrack {11},\text{ Theorem 2.9 }\right\rbrack } \) Let \( a \in R \) . Then \( a \in {R}^{\text{® }} \) with \( I\left( a\right) = m \) if and only if \( a \in {\mathbb{R}}^{D} \) with ind \( \left( a\right) = m \) and \( {c}_{a} \in {\mathbb{R}}^{ \circledast }. \) In this case, \( {a}^{\cir...
Proof Supposing \( a \in {R}^{ \odot } \) with \( I\left( a\right) = m \), we have \( a \in {R}^{D} \) with \( \operatorname{ind}\left( a\right) = m \) by Theorem 9.5 and\n\n\[ \n{a}^{\circledR }{c}_{a}^{2} = {a}^{\circledR }{\left( a{a}^{D}a\right) }^{2} = {a}^{\circledR }{a}^{3}{a}^{D} = \n\]\n\n\[ \n{a}^{\circledR }...
Yes
Theorem 9.12. \( {}^{\left\lbrack {11},\text{ Theorem }{2.10}\right\rbrack } \) Let \( a, x \in R \) . Then the following conditions are equivalent:\n\n(1) \( {a}^{ \odot } = x \) .\n\n(2) \( {ax} = x \), and \( {xR} = {x}^{ * }R = {a}^{m}R \) for some positive integer \( m \) .\n\n(3) \( {xax} = x,{xR} = {a}^{m}R \) a...
Proof (1) \( \Rightarrow \) (2). Suppose \( {a}^{ \odot } = x \) with \( I\left( a\right) = m \) . Then by the definition of pseudo core inverse, we have \( {xax} = x \).\n\n\[ \n{xR} \subseteq {x}^{ * }R\text{ since }x = a{x}^{2} = \left( {ax}\right) x = {\left( ax\right) }^{ * }x = {x}^{ * }{a}^{ * }x \in {x}^{ * }R;...
Yes
Theorem 9.13. \( {}^{\left\lbrack {11},\text{ Theorem 2.12 }\right\rbrack } \) Let \( a \in R \) . Then \( a \in {R}^{\text{(C) }} \) if and only if there exist \( u, v \in R \) and positive integer \( p, q \) such that\n\n\[ \n{a}^{p} = u{\left( {a}^{ * }\right) }^{p + 1}{a}^{p},{a}^{q} = v{a}^{q + 1}.\n\]\n
Proof Suppose \( {a}^{ \odot } = x \) with \( I\left( a\right) = m \), then\n\n\[ \nx{a}^{m + 1} = {a}^{m}, a{x}^{2} = x,{\left( ax\right) }^{ * } = {ax}.\n\]\n\nBy Lemma 9.3, we have \( {a}^{m}{x}^{m}{a}^{m} = {a}^{m},\;({a}^{m}{x}^{m}{)}^{ \ast } = {a}^{m}{x}^{m} \) which yields \( {a}^{m} = ({a}^{m}{x}^{m}{)}^{ \ast...
Yes
Proposition 9.15. \( {}^{\left\lbrack {11},\text{ Proposition }{2.14}\right\rbrack } \) Let \( a, x \in R \) . Then \( \left( 1\right) \Leftrightarrow \left( 2\right) ,\left( 3\right) \Rightarrow \left( 1\right) \) , where\n\n(1) \( {a}^{ \circledast } = x \) .\n\n(2) \( x{a}^{m + 1} = {a}^{m}, a{x}^{2} = x,{\left( ax\...
Proof \( \left( 1\right) \Rightarrow \left( 2\right) \) . It is clear.\n\n\( \left( 2\right) \Rightarrow \left( 1\right) \) . By Theorem 9.3, we have\n\n\[ x = {xax} = x{a}^{m}{x}^{m} = x\left( {x{a}^{m + 1}}\right) {x}^{m} = {x}^{2}{a}^{m + 1}{x}^{m}, \]\n\nthen \( x = {x}^{m + 1}{a}^{2m}{x}^{m} \) by induction. So \(...
Yes
Theorem 9.18. \( {}^{\left\lbrack {37},\text{ Theorem }{3.3}\right\rbrack } \) Let \( a \in R \) and \( k \in {\mathbb{N}}^{ + } \) . Then the following conditions are equivalent:\n\n(1) \( a \in {R}^{ \odot } \) with \( I\left( a\right) \leq k \) .\n\n(2) There exists a projection \( p \in R \) such that \( {pR} = {a}...
Proof \( \;\left( 1\right) \Rightarrow \left( 2\right) .\; \) Assume that \( a \in \mathbb{R} \) is pseudo core invertible and \( I\left( a\right) \leq k. \) Set \( p = a{a}^{\circledR } \) . By Lemma 9.3 (1), we have \( p = a{a}^{\circledR } = {a}^{k}{\left( {a}^{\circledR }\right) }^{k} = {a}^{k + 1}{\left( {a}^{\cir...
Yes
Theorem 9.20. \( {}^{\left\lbrack {37},\text{ Theorem }{3.5}\right\rbrack } \) Let \( a \in {R}^{\circledR } \) with \( I\left( a\right) \leq k \) . Then \( {a}^{\circledR } \) is EP and \( {\left( {a}^{\circledR }\right) }^{ \dagger } = {\left( {a}^{\circledR }\right) }^{\# } = {a}^{k + 1}{\left( {a}^{k}\right) }^{\le...
Proof Suppose that \( x = {a}^{k + 1}{\left( {a}^{k}\right) }^{\left( 1,3\right) } \) . From Theorem 9.5, we know that \( {a}^{\circledR } = \) \( {a}^{D}{a}^{k}{\left( {a}^{k}\right) }^{\left( 1,3\right) } \) . Then we have\n\n\[ \n{a}^{\circledR }x{a}^{\circledR } = {a}^{\circledR }{a}^{k + 1}{\left( {a}^{k}\right) }...
Yes
Theorem 9.24. \( {}^{\left\lbrack {37},\text{ Theorem }{3.9}\right\rbrack } \) Let \( a \in {R}^{\circledR } \) with \( I\left( a\right) = k \) . Then the following conditions are equivalent:\n\n(1) \( {a}^{k} \) is EP.\n\n(2) There exists a unit \( u \in R \) such that \( {a}^{\circledR } = u{a}^{k} \) .\n\n(3) There ...
Proof \( \;\left( 1\right) \Rightarrow \left( 2\right) . \) Suppose that \( {a}^{k} \) is EP. We know that \( a \in {\mathbb{R}}^{\bigodot } \) with \( I\left( a\right) = k. \) By Theorem 9.5 and [4, Lemma 11.6], we have\n\n\[ \n{a}^{\circledR } = {a}^{D}{a}^{k}{\left( {a}^{k}\right) }^{\left( 1,3\right) } = {a}^{D}{a}...
Yes
Proposition 10.1. \( {}^{\left\lbrack {11},\text{ Proposition }{4.2}\right\rbrack } \) Let \( a, x \in R \) with \( {ax} = {xa},{a}^{ * }x = x{a}^{ * } \) . If \( a \in {R}^{ \odot } \), then \( {a}^{ \odot }x = x{a}^{ \odot } \) .
Proof Supoose that \( \operatorname{ind}\left( a\right) = m \) . From the condition \( {ax} = {xa},{a}^{ * }x = x{a}^{ * } \), we have \( {a}^{m}x = x{a}^{m},{\left( {a}^{m}\right) }^{ * }x = x{\left( {a}^{m}\right) }^{ * } \) . Then\n\n\[ \n{a}^{\circledR }x = {a}^{\circledR }a{a}^{\circledR }x = {a}^{\circledR }{a}^{...
Yes
Theorem 10.2. \( {}^{\left\lbrack {11},\text{ Theorem }{4.3}\right\rbrack }\; \) Let \( a, b \in {R}^{\text{® }} \) with \( {ab} = {ba} \) and \( a{b}^{ * } = {b}^{ * }a \) . Then \( {\left( ab\right) }^{ \odot } = {a}^{ \odot }{b}^{ \odot } = {b}^{ \odot }{a}^{ \odot }.
Proof From Proposition 10.1, it follows that\n\n\[ \n{b}^{\circledR }a = a{b}^{\circledR }\text{ and }{a}^{\circledR }b = b{a}^{\circledR }. \]\n\nThe condition \( {b}^{ * }a = a{b}^{ * },{a}^{ * }{b}^{ * } = {b}^{ * }{a}^{ * } \) ensures that \( {b}^{ * }{a}^{\circledR } = {a}^{\circledR }{b}^{ * } \), which together ...
Yes
Theorem 11.1. \( {}^{\left\lbrack {30},\text{ Theorem }{3.3}\right\rbrack } \) If \( a \in {R}^{D} \), then the following conditions are equivalent:\n\n(1) \( a \in {R}^{ \odot } \) .\n\n(2) \( a{a}^{D} \in {R}^{\{ 1,3\} } \) .\n\n(3) \( {a}^{\pi } \in {R}^{\{ 1,4\} } \) .\n\nIn this case, \( a{a}^{\circledR } \in \lef...
Proof (1) \( \Rightarrow \) (2). Suppose that \( I\left( a\right) = m \) . Then we have\n\n\[ \n{a}^{D}a{a}^{\circledR } = {a}^{D}{a}^{m + 1}{\left( {a}^{\circledR }\right) }^{m + 1} = {a}^{m}{\left( {a}^{\circledR }\right) }^{m + 1} = {a}^{\circledR },\n\]\n\n(11.71)\n\nand\n\n\[ \n{a}^{\circledR }a{a}^{D} = {a}^{\cir...
Yes
Theorem 11.2. \( {}^{\left\lbrack {30},\text{ Theorem }{3.4}\right\rbrack } \) If \( a \in {R}^{D} \), then the following conditions are equivalent:\n\n(1) \( a \in {R}^{ \oplus } \).\n\n(2) \( a{a}^{D} \in {R}^{\{ 1,4\} } \).\n\n(3) \( {a}^{\pi } \in {R}^{\{ 1,3\} } \).\n\nIn this case, \( {a}^{\circledR }a \in \left(...
\[ {a}_{\circledR } = {\left( a{a}^{D}\right) }^{\left( 1,4\right) }{a}^{D} = \left( {1 - {a}^{\pi }{\left( {a}^{\pi }\right) }^{\left( 1,3\right) }}\right) {a}^{D}, \] for any \( {\left( a{a}^{D}\right) }^{\left( 1,4\right) } \in \left( {a{a}^{D}}\right) \{ 1,4\} \) and \( {\left( {a}^{\pi }\right) }^{\left( 1,3\right...
Yes
Lemma 11.4. \( {}^{\left\lbrack 4,\text{ Theorem 11.4 }\right\rbrack } \) Let \( {a}_{1},{a}_{2}, d \in R \) . If \( {a}_{1},{a}_{2} \in {R}^{D} \) and \( d{a}_{1} = {a}_{2}d \) , then \( {a}_{2}^{D}d = d{a}_{1}^{D} \) .
In fact, noting that \( {\alpha a} = {a\beta } \), we have \( {\alpha }^{D}a = a{\beta }^{D} \) by Lemma 11.4. Thus,\n\n\[ {\beta }^{D} = \left( {\beta {\beta }^{D} + {\beta }^{D} - \beta {\beta }^{D}}\right) \]\n\n\[ = \left( {\beta {\beta }^{D} + \left( {1 - \beta }\right) {\beta }^{D}}\right) \]\n\n\[ = \left( {\bet...
No
Theorem 11.6. \( {}^{\left\lbrack {30},\text{ Theorem }{3.10}\right\rbrack } \) Let \( a, b \in R \) . If \( \alpha = 1 - {ab} \in {R}^{ \odot } \), then the following conditions are equivalent:\n\n(1) \( \beta = 1 - {ba} \in {R}^{ \odot } \) .\n\n(2) \( b{\alpha }^{\pi }{ra} \in {R}^{\{ 1,4\} } \) .\n\n(3) \( v = \lef...
Proof Since \( \alpha = 1 - {ab} \in {R}^{ \odot } \), suppose that \( I\left( \alpha \right) = k \), it follows that \( \alpha \in {R}^{D} \) with \( i\left( \alpha \right) = k \) by Theorem 9.5 and \( {\alpha }^{\pi } \in {R}^{\{ 1,4\} } \) with \( 1 - \alpha {\alpha }^{\circledR } \in \left( {\alpha }^{\pi }\right) ...
Yes
Lemma 13.4 \( {}^{\left\lbrack 5,\text{ Lemma 3.1 }\right\rbrack } \) If \( a \in R \) with \( {a}^{\circledR } \) exists and \( j \in J\left( R\right) \), then \( 1 - \sigma \) is invertible.
Proof Since\n\n\[ \sigma = \alpha {a}^{\circledR }a{\alpha }^{-1}\left( {1 - a{a}^{\circledR }}\right) \beta \]\n\n\[ = \alpha {a}^{\circledR }a\left( {1 + {a}^{\circledR }j}\right) \left( {1 - a{a}^{\circledR }}\right) \beta \]\n\n\[ = \alpha {a}^{\circledR }a\left( {1 - a{a}^{\circledR }}\right) \beta + \alpha {a}^{\...
Yes
Theorem 13.5 \( {}^{\left\lbrack 5,\text{ Theorem 3.2}\right\rbrack } \) Suppose that \( a \in R \) with \( {a}^{\circledR } \) exists and \( j \in J\left( R\right) \) . Let \( f = a + j - \varepsilon \) . If \( j\left( {{a}^{\circledR }a - 1}\right) a = 0 \) and \( a\left( {{a}^{\circledR }a - 1}\right) j = 0 \), then...
Proof By Lemma 13.4, we know that \( 1 - \sigma \) is invertible. Since \( j \in J\left( R\right) \), we obtain that \( \gamma = \alpha \left( {1 - {a}^{\circledR }a}\right) {ja}@\beta \in J\left( R\right) \) . So, we have \( 1 - \gamma \) is invertible. Since \( {\beta }^{ * }\left( {1 - a{a}^{\circledR }}\right) j{a}...
Yes
Proposition 14. \( {1}^{\left\lbrack {12},\text{ Proposition }1\right\rbrack } \) Let \( A \in {R}^{m \times n} \) with \( A\{ 1,4\} \neq \varnothing \), and \( T \) a square matrix with \( {T}^{k} = {PAQ} \) for some positive integer \( k \) and matrices \( P \) and \( Q \) . Suppose \( {P}^{\prime }{PA} = A = {AQ}{Q}...
Proof Applying [4, Theorem 11.6], \( {T}^{D} \) exists if and only if \( {A}^{\left( 1,4\right) }{AQTPA} + I - \) \( {A}^{\left( 1,4\right) }A \) is invertible, in which case, \( {T}^{D} = {PA}{\left\lbrack {A}^{\left( 1,4\right) }AQTPA + I - {A}^{\left( 1,4\right) }A\right\rbrack }^{-1}Q \) .\n\nIf \( A\{ 1,4\} \neq \...
Yes
Theorem 14.3 \( {}^{\left\lbrack {12},\text{ Proposition }3\right\rbrack }\; \) Let \( A \) be regular with \( {A}^{ - } \in A\{ 1\} \), and \( T \) a square matrix with \( {T}^{k} = {PAQ} \) for some positive integer \( k \) and matrices \( P, Q \) . Suppose\n\n\[ \nU = {A}^{ - }{AQTPA} + I - {A}^{ - }A, V = {AQ}{\lef...
Proof By [4, Theorem 16.1], \( {T}^{D} \) exists with ind \( \left( T\right) \leq k, A = {P}^{\prime }{PA} = {AQ}{Q}^{\prime } \) for some matrices \( {P}^{\prime } \) and \( {Q}^{\prime } \) if and only if \( U \) is invertible. Next, we prove that \( U \) is invertible if and only if \( {XAQTPA} = A = {AQTPAY} \) for...
Yes
Theorem 14.5 Let \( A \) be a matrix with \( A\{ 1,3\} \neq \varnothing \) and \( T \) a square matrix with \( {T}^{k} = {PAQ} \) for some positive integer \( k \), matrices \( P \) and \( Q. \) Suppose \( {P}^{\prime }{PA} = A = {AQ}{Q}^{\prime } \) for some matrix \( {P}^{\prime } \) and \( {Q}^{\prime } \), and\n\n\...
Proof Applying Theorem 9.5, \( {T}^{\circledR } \) exists if and only if \( {T}^{D} \) exists and \( {T}^{k}\{ 1,3\} \neq \varnothing \) .\n\nSince \( {P}^{\prime }{PA} = A = {AQ}{Q}^{\prime } \), we get \( {T}^{D} \) exists with \( \operatorname{ind}\left( T\right) \leq k \) if and only if\n\n\( {A}^{\left( 1,3\right)...
Yes
Corollary 14.6 Let \( A \) be a matrix with \( A\{ 1,3\} \neq \varnothing \) and \( T \) a square matrix with \( T = {PA} \) for some matrix \( P \) . Suppose \( {P}^{\prime }{PA} = A \) for some matrix \( {P}^{\prime } \) . Then the following are equivalent:\n\n(1) \( {T}^{ \circledast } \) exists.\n\n(2) \( U \) is i...
The following result characterizes the dual pseudo core invertibility of \( T \) .
No
Theorem 15. \( {\mathbf{1}}^{\left\lbrack {12},\text{Theorem }1\right\rbrack }\; \) Let \( T = \left\lbrack \begin{array}{ll} 0 & 0 \\ b & d \end{array}\right\rbrack \in {R}^{2 \times 2} \) . Then \( {T}^{\text{® }} \) exists if and only if \( {d}^{ \oplus } \) exists. Moreover, \( {T}^{ \oplus } = \left\lbrack \begin{...
Proof Suppose \( {d}^{ \odot } \) exists with \( I\left( d\right) = k \), then it is easy to check that \( {T}^{ \odot } \) exists\n\nwith \( {T}^{\left( \Theta \right) } = \left\lbrack \begin{matrix} 0 & 0 \\ 0 & {d}^{\left( \Theta \right) } \end{matrix}\right\rbrack \), as well as\n\n\[ I\left( T\right) = \left\{ \be...
Yes
Theorem 15. \( {2}^{\left\lbrack {12},\text{ Theorem }2\right\rbrack } \) Let \( T = \left\lbrack \begin{array}{ll} a & 0 \\ b & d \end{array}\right\rbrack \in {R}^{2 \times 2} \) such that \( d \in {R}^{\circledR } \cap {R}_{\circledR } \) with \( \;I\left( d\right) = k,\;{a}^{k}\{ 1,4\} \neq \varnothing ,\;c = \matho...
Proof Applying Theorem 9.14, \( d \in {R}^{\circledR } \cap {R}_{\circledR } \) with \( I\left( d\right) = k \) if and only if \( {d}^{D} \) and \( {\left( {d}^{k}\right) }^{ \dagger } \) exist. In this case, \( {d}^{k}{\left( {d}^{k}\right) }^{ \dagger } = {d}^{k}{\left( {d}^{k}\right) }^{ \odot },\left( {d}^{k + 1}\r...
No
Theorem 15. \( {3}^{\left\lbrack {12},\text{Theorem 3}\right\rbrack } \) Let \( T = \left\lbrack \begin{array}{ll} a & 0 \\ b & d \end{array}\right\rbrack \in {R}^{2 \times 2} \) such that \( a \in {R}^{ \odot } \) with \( I\left( a\right) = k,{d}^{k}\{ 1,3\} \neq \varnothing, c = \mathop{\sum }\limits_{{i + j = k - 1}...
Proof Since \( T = \left\lbrack \begin{array}{ll} a & 0 \\ b & d \end{array}\right\rbrack \), then\n\n\[ {T}^{k} = {\left\lbrack \begin{array}{ll} a & 0 \\ b & d \end{array}\right\rbrack }^{k} = \left\lbrack \begin{matrix} {a}^{k} & 0 \\ c & {d}^{k} \end{matrix}\right\rbrack \]\n\n
Yes
Lemma 1 Let \( Z = \mathop{\sum }\limits_{{\alpha \in {G}_{4, n}}}p\left( \alpha \right) {e}_{\alpha }^{ \bullet } \) be decomposable, and \( p\left( {1,1,1,1}\right) = 1 \), its spanning matrix is\n\n\[ A = \left( \begin{matrix} 1 & {a}_{12} & \cdots & {a}_{1i} & \cdots & {a}_{1j} & \cdots & {a}_{1n} \\ 1 & {a}_{22} &...
Proof In order to prove \( B = \left( \begin{array}{l} {\alpha }_{1} \\ {\alpha }_{2} \\ {\alpha }_{3} \end{array}\right) \) is spanning matrix of decomposable element\n\n\( W = \mathop{\sum }\limits_{{\beta = \left( {i, j, k}\right) \in {G}_{3, n}}}q\left( \beta \right) {e}_{\beta }^{ \bullet } \) in \( {V}^{\left( 3\...
Yes
Theorem 1 Let \( Z = \mathop{\sum }\limits_{{\alpha \in {G}_{4, n}}}p\left( \alpha \right) {e}_{\alpha }^{ \bullet } \) and \( A \) are similar with lemma 1. If \( {f}_{2}\left( x\right) \) (without less of generality) has one simple root and one triple root, and assume \( {a}_{12} = \) \( {a}_{22} = {a}_{32} = a,{a}_{...
Proof By assumption, \( A \) is spanning matrix of decomposable element \( Z = \mathop{\sum }\limits_{{\alpha \in {G}_{4, n}}}p\left( \alpha \right) {e}_{\alpha }^{ \bullet } \), where \n\n\[ \nA = \left( \begin{matrix} 1 & a & \cdots & {a}_{1i} & \cdots & {a}_{1j} & \cdots & {a}_{1n} \\ 1 & a & \cdots & {a}_{2i} & \cd...
Yes
Theorem 3 Let \( Z = \mathop{\sum }\limits_{{\alpha \in {G}_{4, n}}}p\left( \alpha \right) {e}_{\alpha }^{ \bullet } \) and \( A \) are similar with lemma 1 . If \( {f}_{2}\left( x\right) \) has two double roots, assume \( {a}_{12} = {a}_{22} = a,{a}_{32} = {a}_{42} = b\left( {b \neq a}\right) \), then sub-matrix \( B ...
Proof In order to prove \( {W}_{1} = \mathop{\sum }\limits_{{\beta \in {G}_{2, n}}}S\left( \beta \right) {e}_{\beta }^{ \bullet } \) is decomposable, and \( B = \left( \begin{array}{l} {\alpha }_{1} \\ {\alpha }_{2} \end{array}\right) \) is spanning matrix of \( {W}_{1} \), we need only to prove that we always have\n\n...
No
Lemma 7 Suppose that the condition \( \left( \mathbf{K}\right) \) holds, \( T \in \left( {0,\infty }\right) ,\bar{g},\bar{h} \in {C}^{1}\left( \left\lbrack {-\tau, T}\right\rbrack \right) ,\bar{u} \in \) \( C\left( {\bar{\mho }}_{T}\right) \cap {C}^{1,2}\left( {\mho }_{T}\right) \) satisfies \( \bar{u} \leq {u}^{ * } \...
By the condition \( \left( \mathbf{K}\right) \), a direct computation shows that for \( \left( {t, x}\right) \in {\mathcal{O}}_{{t}^{ * }} \), there exist \( {\xi }_{1} \) between \( \bar{u}\left( {t, x}\right) ,{u}_{\varepsilon }\left( {t, x}\right) \) and \( {\xi }_{2} \) between \( \bar{u}\left( {t - \tau, x}\right)...
Yes
Lemma 8 Suppose that the condition \( \left( \mathbf{K}\right) \) holds, \( T \in \left( {0,\infty }\right) ,\bar{g},\bar{h} \in {C}^{1}\left( \left\lbrack {-\tau, T}\right\rbrack \right) ,\bar{u} \in \) \( C\left( {\bar{\mho }}_{T}\right) \cap {C}^{1,2}\left( {\mho }_{T}\right) \) with \( {\mho }_{T} = \left\{ {\left(...
A minor modification of the proof of Lemma 7 yields Lemma 8.
No
Lemma 10 Assume that the condition \( \left( \mathbf{K}\right) \) holds. Let \( \left( {u, g, h}\right) \) be a solution of the problem (1.6). Then the following three assertions are equivalent:\n\n(i) \( {h}_{\infty } \) or \( {g}_{\infty } \) is finite;\n\n(ii) \( {h}_{\infty } - {g}_{\infty } \leq {R}^{ * } \) ;\n\n...
Proof The proof of this lemma follows that of Lemma 2.2 in [3].
No
Lemma 11 Assume that \( \left( \mathbf{K}\right) \) holds. Let \( \left( {u, g, h}\right) \) be a solution of (1.6). Then vanishing happens provided that \( h\left( 0\right) - g\left( 0\right) < {R}^{ * } \) and \( \parallel \phi {\parallel }_{{L}^{\infty }\left( {\left\lbrack {-\tau ,0}\right\rbrack \times \left\lbrac...
Proof Set\n\n\[ \n{h}_{0} = \frac{h\left( 0\right) - g\left( 0\right) }{2}. \]\n\nFor any \( \varepsilon > 0 \), we can find a small positive constant \( \delta \) such that\n\n\[ \n{\pi \mu \delta } \leq {\varepsilon }^{2}{h}_{0}^{2} \]\n\n\[ \nf\left( {u\left( {t, x}\right), u\left( {t - \tau, x}\right) }\right) \leq...
Yes
Example 3.1 Let \( f\left( {x, u}\right) = {\lambda h}\left( x\right) {u}^{q - 1} \) in (1.1), where \( \lambda, q \in \mathbb{R},0 < h\left( x\right) \in {C}^{1}\left( \bar{\Omega }\right) \) and condition \( \left( V\right) \) holds, then (1.1) has no any positive solutions under assumptions in Theorem 2.1 and the co...
In fact, since \( F\left( {x,0}\right) = 0 \), we can conclude that\n\n\[ F\left( {x, u}\right) = {\int }_{0}^{u}{\lambda h}\left( x\right) {t}^{q - 1}\mathrm{\;d}t = \frac{\lambda }{q}h\left( x\right) {u}^{q} \]\n\nand\n\n\[ {F}_{1}\left( {x, u}\right) = \frac{\lambda }{q}{u}^{q}\mathop{\sum }\limits_{{i = 1}}^{n}{V}_...
Yes
For all integers \( n \geq 1 \), let the sequence \( {P}_{n} \) be defined by (1.7). Then\n\n\[\frac{1}{{180}{\left( n + \alpha \right) }^{4}} \leq \gamma - {P}_{n} < \frac{1}{{180}{\left( n + \beta \right) }^{4}}\]\n\n(3.1)\n\nwith the best possible constants\n\n\[\alpha = \frac{1}{\sqrt[4]{{90}\ln \left( {7/3}\right)...
Proof The inequality (3.1) can be written as\n\n\[\alpha \geq \frac{1}{\sqrt[4]{{180}\left( {\frac{1}{2}\ln \left( {{n}^{2} + n + \frac{1}{3}}\right) - \psi \left( {n + 1}\right) }\right) }} - n > \beta .\]\n\nIn order to prove (3.1), we define the function \( M\left( x\right) \) by\n\n\[M\left( x\right) = {\left( u\le...
Yes
Theorem 2 The harmonic number has the following asymptotic expansion:\n\n\[ \n{H}_{n} \sim \frac{1}{2}\ln \left( {{n}^{2} + n + \frac{1}{3}}\right) + \gamma + \mathop{\sum }\limits_{{j = 2}}^{\infty }\frac{{a}_{j}}{{\left( n + \frac{1}{2}\right) }^{2j}} \n\]
Proof Let \( m = \frac{1}{2}n\left( {n + 1}\right) \) . The asymptotic expansion (3.4) can be written as\n\n\[ \n{H}_{n} \sim \frac{1}{2}\ln \left( {2m}\right) + \frac{1}{2}\ln \left( {1 + \frac{1}{6m}}\right) + \gamma + \mathop{\sum }\limits_{{j = 2}}^{\infty }\frac{{a}_{j}}{{\left( 2m\right) }^{j}}{\left( 1 + \frac{1...
Yes
Theorem 3 Let \( {P}_{n} \) be defined by (1.7). Then\n\n\[ \frac{\frac{1}{180}}{{\left( n + \frac{1}{2}\right) }^{4}} - \frac{\frac{17}{4536}}{{\left( n + \frac{1}{2}\right) }^{6}} + \frac{\frac{107}{25920}}{{\left( n + \frac{1}{2}\right) }^{8}} - \frac{\frac{6467}{855360}}{{\left( n + \frac{1}{2}\right) }^{10}} \]\n\...
Proof Here we only prove the first inequality in (3.18). The proof of the second inequality in (3.18) is similar, we omit them. The lower bound of (3.18) is obtained by considering the function \( G\left( x\right) \) for \( x \geq 1 \) defined by\n\n\[ G\left( x\right) = \psi \left( {x + 1}\right) - \frac{1}{2}\ln \lef...
Yes
Theorem 2 The period of \( {D}_{p} \) is\n\n\[ \n{T}_{p} = \left( {2 - 1}\right) \left( {3 - 1}\right) \left( {5 - 1}\right) \cdots \left( {p - 1}\right) .\n\]
Proof We prove this by induction. The case for \( {D}_{3} \) has been proved. For simplicity, we try to explain the procedure by proving the case for \( {D}_{5} \) . First, we observe that\n\n\[ \n{M}_{{p}_{n + 1}} = {M}_{{p}_{n}} - \left\{ {{p}_{n + 1}h \mid h \in {M}_{{p}_{n}}}\right\} .\n\]\n\nWe show this procedure...
No