|
|
Line 1: |
Line 1: |
− | An integer <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d0324801.png" />, defined for every topological space <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d0324802.png" /> of a given class <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d0324803.png" />, which has sufficiently many properties to make it resemble the usual notion of dimension: the number of coordinates of higher-dimensional Euclidean spaces. Here one requires of the class <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d0324804.png" /> that it contains all cubes with any number of coordinates, and together with any space <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d0324805.png" /> which is an element of it, it should also contain as an element every space homeomorphic to <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d0324806.png" />. For a dimension invariant <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d0324807.png" /> it is assumed, in any case, that for homeomorphic spaces <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d0324808.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d0324809.png" /> one always has <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248010.png" />, and that for the <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248011.png" />-dimensional cube <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248012.png" /> one has <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248013.png" />. Among the dimension invariants, the most important ones are the so-called classical dimensions — the [[Lebesgue dimension|Lebesgue dimension]] <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248014.png" /> and the (large and small) inductive dimensions (cf. [[Inductive dimension|Inductive dimension]]) <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248015.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248016.png" />.
| + | <!-- |
| + | d0324801.png |
| + | $#A+1 = 83 n = 0 |
| + | $#C+1 = 83 : ~/encyclopedia/old_files/data/D032/D.0302480 Dimension invariant |
| + | Automatically converted into TeX, above some diagnostics. |
| + | Please remove this comment and the {{TEX|auto}} line below, |
| + | if TeX found to be correct. |
| + | --> |
| | | |
− | The following propositions distinguish <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248017.png" /> from all other dimension invariants defined, respectively, in the class of all (metric) compacta, all metrizable and all separable metrizable spaces, and hence settle for these spaces the problem of the axiomatic definition of dimension. The only dimension invariant <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248018.png" /> satisfying the conditions 1), 2), 3) listed below and defined in the class <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248019.png" /> of all (metric) compacta <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248020.png" /> is the dimension <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248021.png" /> (Aleksandrov's theorem).
| + | {{TEX|auto}} |
| + | {{TEX|done}} |
| | | |
− | Condition 1) (Poincaré's axiom). If a space <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248022.png" /> is of class <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248023.png" /> and if <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248024.png" /> is equal to the non-negative integer <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248025.png" />, then <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248026.png" /> contains a closed subspace <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248027.png" /> for which <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248028.png" /> and such that the set <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248029.png" /> is disconnected.
| + | An integer $ d ( X) $, |
| + | defined for every topological space $ X $ |
| + | of a given class $ {\mathcal K} $, |
| + | which has sufficiently many properties to make it resemble the usual notion of dimension: the number of coordinates of higher-dimensional Euclidean spaces. Here one requires of the class $ {\mathcal K} $ |
| + | that it contains all cubes with any number of coordinates, and together with any space $ X $ |
| + | which is an element of it, it should also contain as an element every space homeomorphic to $ X $. |
| + | For a dimension invariant $ d ( X) $ |
| + | it is assumed, in any case, that for homeomorphic spaces $ X $ |
| + | and $ X ^ { \prime } $ |
| + | one always has $ d ( X) = d ( X ^ { \prime } ) $, |
| + | and that for the $ n $- |
| + | dimensional cube $ I ^ { n } $ |
| + | one has $ d ( I ^ { n } ) = n $. |
| + | Among the dimension invariants, the most important ones are the so-called classical dimensions — the [[Lebesgue dimension|Lebesgue dimension]] $ \mathop{\rm dim} X $ |
| + | and the (large and small) inductive dimensions (cf. [[Inductive dimension|Inductive dimension]]) $ \mathop{\rm Ind} X $, |
| + | $ \mathop{\rm ind} X $. |
| | | |
− | Condition 2) (the finite sum axiom). If a space <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248030.png" /> of class <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248031.png" /> is the union of two closed subspaces <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248032.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248033.png" /> for which <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248034.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248035.png" />, then also <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248036.png" />.
| + | The following propositions distinguish $ \mathop{\rm dim} X $ |
| + | from all other dimension invariants defined, respectively, in the class of all (metric) compacta, all metrizable and all separable metrizable spaces, and hence settle for these spaces the problem of the axiomatic definition of dimension. The only dimension invariant $ d ( X) $ |
| + | satisfying the conditions 1), 2), 3) listed below and defined in the class $ {\mathcal K} $ |
| + | of all (metric) compacta $ X $ |
| + | is the dimension $ \mathop{\rm dim} X = \mathop{\rm Ind} X = \mathop{\rm ind} X $( |
| + | Aleksandrov's theorem). |
| | | |
− | Condition 3) (Brouwer's axiom in metric form). If <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248037.png" /> is a space belonging to <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248038.png" /> and if <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248039.png" /> is the non-negative integer <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248040.png" />, then there is a positive number <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248041.png" /> such that for every space <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248042.png" /> which is the image of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248043.png" /> under some <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248044.png" />-mapping one has the inequality <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248045.png" />. Here a mapping <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248046.png" /> from a compactum <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248047.png" /> onto a compactum <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248048.png" /> is called an <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248050.png" />-mapping if it is continuous and if the complete pre-image <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248051.png" /> of every point <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248052.png" /> has diameter <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248053.png" /> in <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248054.png" />. | + | Condition 1) (Poincaré's axiom). If a space $ X $ |
| + | is of class $ {\mathcal K} $ |
| + | and if $ d ( X) $ |
| + | is equal to the non-negative integer $ n $, |
| + | then $ X $ |
| + | contains a closed subspace $ X _ {0} $ |
| + | for which $ d ( X _ {0} ) < n $ |
| + | and such that the set $ X \setminus X _ {0} $ |
| + | is disconnected. |
| | | |
− | Shchepin's theorem [[#References|[2]]]. The dimension <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248055.png" /> is the only dimension invariant <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248056.png" /> defined, respectively, in the class <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248057.png" /> of all metric, or all separable metric spaces <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248058.png" />, which satisfies the following conditions (Shchepin's theorem): | + | Condition 2) (the finite sum axiom). If a space $ X $ |
| + | of class $ {\mathcal K} $ |
| + | is the union of two closed subspaces $ X _ {1} $ |
| + | and $ X _ {2} $ |
| + | for which $ d ( X _ {1} ) \leq n $, |
| + | $ d ( X _ {2} ) \leq n $, |
| + | then also $ d ( X) \leq n $. |
| + | |
| + | Condition 3) (Brouwer's axiom in metric form). If $ X $ |
| + | is a space belonging to $ {\mathcal K} $ |
| + | and if $ d ( X) $ |
| + | is the non-negative integer $ n $, |
| + | then there is a positive number $ \epsilon $ |
| + | such that for every space $ Y $ |
| + | which is the image of $ X $ |
| + | under some $ \epsilon $- |
| + | mapping one has the inequality $ d ( Y) \geq n $. |
| + | Here a mapping $ f $ |
| + | from a compactum $ X $ |
| + | onto a compactum $ Y $ |
| + | is called an $ \epsilon $- |
| + | mapping if it is continuous and if the complete pre-image $ f ^ {-} 1 ( y) $ |
| + | of every point $ y \in Y $ |
| + | has diameter $ < \epsilon $ |
| + | in $ X $. |
| + | |
| + | Shchepin's theorem [[#References|[2]]]. The dimension $ \mathop{\rm dim} X $ |
| + | is the only dimension invariant $ d ( X) $ |
| + | defined, respectively, in the class $ {\mathcal K} $ |
| + | of all metric, or all separable metric spaces $ X $, |
| + | which satisfies the following conditions (Shchepin's theorem): |
| | | |
| Condition 1) (Poincaré's axiom). See above. | | Condition 1) (Poincaré's axiom). See above. |
| | | |
− | Condition 2) (the countable sum axiom). If a space <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248059.png" /> belonging to the class <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248060.png" /> is the union of a countable number of closed subspaces <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248061.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248062.png" /> each having <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248063.png" />, then also <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248064.png" />. | + | Condition 2) (the countable sum axiom). If a space $ X $ |
| + | belonging to the class $ {\mathcal K} $ |
| + | is the union of a countable number of closed subspaces $ X _ {k} $, |
| + | $ k = 1, 2 \dots $ |
| + | each having $ d ( X _ {k} ) \leq n $, |
| + | then also $ d ( X) \leq n $. |
| | | |
− | Condition 3) (Brouwer's axiom in general form). If for a space <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248065.png" /> belonging to the class <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248066.png" /> one has <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248067.png" />, then there is a finite open covering <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248068.png" /> of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248069.png" /> such that for every space <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248070.png" /> belonging to <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248071.png" /> and which is the image of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248072.png" /> under some <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248074.png" />-mapping one has <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248075.png" />. Here a mapping <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248076.png" /> from a space <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248077.png" /> on which a certain open covering <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248078.png" /> has been fixed onto a space <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248079.png" /> is called an <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248080.png" />-mapping if every point <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248081.png" /> of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248082.png" /> has a neighbourhood <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248083.png" /> whose complete pre-image <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248084.png" /> is contained in some element of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/d/d032/d032480/d03248085.png" />. | + | Condition 3) (Brouwer's axiom in general form). If for a space $ X $ |
| + | belonging to the class $ {\mathcal K} $ |
| + | one has $ d ( X) \leq n $, |
| + | then there is a finite open covering $ \omega $ |
| + | of $ X $ |
| + | such that for every space $ Y $ |
| + | belonging to $ {\mathcal K} $ |
| + | and which is the image of $ X $ |
| + | under some $ \omega $- |
| + | mapping one has $ d ( Y) \geq n $. |
| + | Here a mapping $ f $ |
| + | from a space $ X $ |
| + | on which a certain open covering $ \omega $ |
| + | has been fixed onto a space $ Y $ |
| + | is called an $ \omega $- |
| + | mapping if every point $ y $ |
| + | of $ Y $ |
| + | has a neighbourhood $ O _ {y} $ |
| + | whose complete pre-image $ f ^ {-} 1 ( O _ {y} ) $ |
| + | is contained in some element of $ \omega $. |
| | | |
| ====References==== | | ====References==== |
| <table><TR><TD valign="top">[1]</TD> <TD valign="top"> P.S. Aleksandrov, "Some old problems in homological dimension theory" , ''Proc. Internat. Symp. Topology and its Applications. Herzog-Novi, 1968'' , Beograd (1969) pp. 38–42 (In Russian)</TD></TR><TR><TD valign="top">[2]</TD> <TD valign="top"> E. [E. Shchepin] Ščepin, "Axiomatics of the dimension of metric spaces" ''Soviet Math. Dokl.'' , '''13''' (1972) pp. 1177–1179 ''Dokl. Akad. Nauk SSSR'' , '''206''' : 1 (1972) pp. 31–32</TD></TR><TR><TD valign="top">[3]</TD> <TD valign="top"> P.S. Aleksandrov, B.A. Pasynkov, "Introduction to dimension theory" , Moscow (1973) (In Russian)</TD></TR></table> | | <table><TR><TD valign="top">[1]</TD> <TD valign="top"> P.S. Aleksandrov, "Some old problems in homological dimension theory" , ''Proc. Internat. Symp. Topology and its Applications. Herzog-Novi, 1968'' , Beograd (1969) pp. 38–42 (In Russian)</TD></TR><TR><TD valign="top">[2]</TD> <TD valign="top"> E. [E. Shchepin] Ščepin, "Axiomatics of the dimension of metric spaces" ''Soviet Math. Dokl.'' , '''13''' (1972) pp. 1177–1179 ''Dokl. Akad. Nauk SSSR'' , '''206''' : 1 (1972) pp. 31–32</TD></TR><TR><TD valign="top">[3]</TD> <TD valign="top"> P.S. Aleksandrov, B.A. Pasynkov, "Introduction to dimension theory" , Moscow (1973) (In Russian)</TD></TR></table> |
− |
| |
− |
| |
| | | |
| ====Comments==== | | ====Comments==== |
An integer $ d ( X) $,
defined for every topological space $ X $
of a given class $ {\mathcal K} $,
which has sufficiently many properties to make it resemble the usual notion of dimension: the number of coordinates of higher-dimensional Euclidean spaces. Here one requires of the class $ {\mathcal K} $
that it contains all cubes with any number of coordinates, and together with any space $ X $
which is an element of it, it should also contain as an element every space homeomorphic to $ X $.
For a dimension invariant $ d ( X) $
it is assumed, in any case, that for homeomorphic spaces $ X $
and $ X ^ { \prime } $
one always has $ d ( X) = d ( X ^ { \prime } ) $,
and that for the $ n $-
dimensional cube $ I ^ { n } $
one has $ d ( I ^ { n } ) = n $.
Among the dimension invariants, the most important ones are the so-called classical dimensions — the Lebesgue dimension $ \mathop{\rm dim} X $
and the (large and small) inductive dimensions (cf. Inductive dimension) $ \mathop{\rm Ind} X $,
$ \mathop{\rm ind} X $.
The following propositions distinguish $ \mathop{\rm dim} X $
from all other dimension invariants defined, respectively, in the class of all (metric) compacta, all metrizable and all separable metrizable spaces, and hence settle for these spaces the problem of the axiomatic definition of dimension. The only dimension invariant $ d ( X) $
satisfying the conditions 1), 2), 3) listed below and defined in the class $ {\mathcal K} $
of all (metric) compacta $ X $
is the dimension $ \mathop{\rm dim} X = \mathop{\rm Ind} X = \mathop{\rm ind} X $(
Aleksandrov's theorem).
Condition 1) (Poincaré's axiom). If a space $ X $
is of class $ {\mathcal K} $
and if $ d ( X) $
is equal to the non-negative integer $ n $,
then $ X $
contains a closed subspace $ X _ {0} $
for which $ d ( X _ {0} ) < n $
and such that the set $ X \setminus X _ {0} $
is disconnected.
Condition 2) (the finite sum axiom). If a space $ X $
of class $ {\mathcal K} $
is the union of two closed subspaces $ X _ {1} $
and $ X _ {2} $
for which $ d ( X _ {1} ) \leq n $,
$ d ( X _ {2} ) \leq n $,
then also $ d ( X) \leq n $.
Condition 3) (Brouwer's axiom in metric form). If $ X $
is a space belonging to $ {\mathcal K} $
and if $ d ( X) $
is the non-negative integer $ n $,
then there is a positive number $ \epsilon $
such that for every space $ Y $
which is the image of $ X $
under some $ \epsilon $-
mapping one has the inequality $ d ( Y) \geq n $.
Here a mapping $ f $
from a compactum $ X $
onto a compactum $ Y $
is called an $ \epsilon $-
mapping if it is continuous and if the complete pre-image $ f ^ {-} 1 ( y) $
of every point $ y \in Y $
has diameter $ < \epsilon $
in $ X $.
Shchepin's theorem [2]. The dimension $ \mathop{\rm dim} X $
is the only dimension invariant $ d ( X) $
defined, respectively, in the class $ {\mathcal K} $
of all metric, or all separable metric spaces $ X $,
which satisfies the following conditions (Shchepin's theorem):
Condition 1) (Poincaré's axiom). See above.
Condition 2) (the countable sum axiom). If a space $ X $
belonging to the class $ {\mathcal K} $
is the union of a countable number of closed subspaces $ X _ {k} $,
$ k = 1, 2 \dots $
each having $ d ( X _ {k} ) \leq n $,
then also $ d ( X) \leq n $.
Condition 3) (Brouwer's axiom in general form). If for a space $ X $
belonging to the class $ {\mathcal K} $
one has $ d ( X) \leq n $,
then there is a finite open covering $ \omega $
of $ X $
such that for every space $ Y $
belonging to $ {\mathcal K} $
and which is the image of $ X $
under some $ \omega $-
mapping one has $ d ( Y) \geq n $.
Here a mapping $ f $
from a space $ X $
on which a certain open covering $ \omega $
has been fixed onto a space $ Y $
is called an $ \omega $-
mapping if every point $ y $
of $ Y $
has a neighbourhood $ O _ {y} $
whose complete pre-image $ f ^ {-} 1 ( O _ {y} ) $
is contained in some element of $ \omega $.
References
[1] | P.S. Aleksandrov, "Some old problems in homological dimension theory" , Proc. Internat. Symp. Topology and its Applications. Herzog-Novi, 1968 , Beograd (1969) pp. 38–42 (In Russian) |
[2] | E. [E. Shchepin] Ščepin, "Axiomatics of the dimension of metric spaces" Soviet Math. Dokl. , 13 (1972) pp. 1177–1179 Dokl. Akad. Nauk SSSR , 206 : 1 (1972) pp. 31–32 |
[3] | P.S. Aleksandrov, B.A. Pasynkov, "Introduction to dimension theory" , Moscow (1973) (In Russian) |
A brief discussion of axioms for dimension invariants can also be found in [a1].
References
[a1] | R. Engelking, "Dimension theory" , North-Holland & PWN (1978) |