reorder vegdist objects

dist_b_reordered <- as.dist(as.matrix(dist_b)[attr(dist_a, "Labels"),

all.equal(dist_a, dist_b_reordered, check.attributes = FALSE)

Dear Gian,
Let's do a small reproducible example:
library(vegan)
data(varespec)
set.seed(12345)
df_a <- t(varespec)
df_b <- t(varespec[, sample.int(ncol(varespec))]) # shuffle the original
column order
dist_a <- vegdist(df_a); dist_b <- vegdist(df_b)
all.equal(dist_a, dist_b, check.attributes = FALSE)

Now, you have two options. If you have access to the original
data.frame/matrix from which the dist object was created, you can
reorder the rows of the second dataset to match that of the first one.
Then do the distance matrix calculation again.
df_b_reordered <- df_b[rownames(df_a), ]
dist_b_reordered1 <- vegdist(df_b_reordered)
all.equal(dist_a, dist_b_reordered1, check.attributes = FALSE)

If you have no access to the original dataset and/or can not recalculate
the distance matrix (e.g. due to computer capacity or time), you can
reorder the distance matrix (both the rows and columns).
dist_b_reordered2 <- as.dist(as.matrix(dist_b)[attr(dist_a, "Labels"),
attr(dist_a, "Labels")])
all.equal(dist_a, dist_b_reordered2, check.attributes = FALSE)

HTH,
?kos Bede-Fazekas
Hungarian Academy of Sciences


2019.06.01. 1:53 keltez?ssel, Gian Maria Niccol? Benucci ?rta:

Hello,

I have two vegan::vegdist objects I want to use into a vegan::mantel

test.

However, rows and columns of the two objects are not in the same order.

Is

there a way to reorder them one another?

For example,

dist_a

         Lycovolu  Hupeaust  Lycofast  Lycodeut
Hupeaust 0.9045557

Lycofast 0.6835441 0.8569781

Lycodeut 0.8036577 0.8294657 0.7958988   ...
...

dist_b

          Lycodeut   Lycofast   Lycoscar   Lycovolu
Lycofast 0.09340659

Lycoscar 0.08974359 0.01831502

Lycovolu 0.09073476 0.01099655 0.01008080   ...
...

Thanks,

Hi Akos,

Thank you for the code. It worked. I reordered the dist objects as I 
cannot access the dataset of one of the two distance matrix.

dist_b_reordered <- as.dist(as.matrix(dist_b)[attr(dist_a, 

"Labels"), attr(dist_a, "Labels")])

all.equal(dist_a, dist_b_reordered, check.attributes = FALSE)

[1] "Mean relative difference: 0.8977224"

There is any statistical test that you suggest to compare the two 
distances? I would like to see if the two matrices show a similar 
pattern, e.g. mantel test?

Thanks,

Gian






On Sat, 1 Jun 2019 at 00:55, Bede-Fazekas ?kos <bfalevlist at gmail.com 
<mailto:bfalevlist at gmail.com>> wrote:

    Dear Gian,
    Let's do a small reproducible example:
    library(vegan)
    data(varespec)
    set.seed(12345)
    df_a <- t(varespec)
    df_b <- t(varespec[, sample.int
    <http://sample.int>(ncol(varespec))]) # shuffle the original
    column order
    dist_a <- vegdist(df_a); dist_b <- vegdist(df_b)
    all.equal(dist_a, dist_b, check.attributes = FALSE)

    Now, you have two options. If you have access to the original
    data.frame/matrix from which the dist object was created, you can
    reorder the rows of the second dataset to match that of the first
    one.
    Then do the distance matrix calculation again.
    df_b_reordered <- df_b[rownames(df_a), ]
    dist_b_reordered1 <- vegdist(df_b_reordered)
    all.equal(dist_a, dist_b_reordered1, check.attributes = FALSE)

    If you have no access to the original dataset and/or can not
    recalculate
    the distance matrix (e.g. due to computer capacity or time), you can
    reorder the distance matrix (both the rows and columns).
    dist_b_reordered2 <- as.dist(as.matrix(dist_b)[attr(dist_a,
    "Labels"),
    attr(dist_a, "Labels")])
    all.equal(dist_a, dist_b_reordered2, check.attributes = FALSE)

    HTH,
    ?kos Bede-Fazekas
    Hungarian Academy of Sciences


    2019.06.01. 1:53 keltez?ssel, Gian Maria Niccol? Benucci ?rta:

    > Hello,
    >
    > I have two vegan::vegdist objects I want to use into a

    vegan::mantel test.

    > However, rows and columns of the two objects are not in the same

    order. Is

    > there a way to reorder them one another?
    >
    > For example,
    >

    >> dist_a

    >? ? ? ? ? Lycovolu? Hupeaust? Lycofast? Lycodeut
    > Hupeaust 0.9045557
    >
    > Lycofast 0.6835441 0.8569781
    >
    > Lycodeut 0.8036577 0.8294657 0.7958988? ?...
    > ...
    >

    >> dist_b

    >? ? ? ? ? ?Lycodeut? ?Lycofast? ?Lycoscar? ?Lycovolu
    > Lycofast 0.09340659
    >
    > Lycoscar 0.08974359 0.01831502
    >
    > Lycovolu 0.09073476 0.01099655 0.01008080? ?...
    > ...
    >
    > Thanks,
    >

Dear Gian,
I'm not a statistician but software developer... hope that other members 
of the mailing list are expert of distance matrices and can answer your 
question.
Have a nice week,
?kos

2019.06.03. 17:44 keltez?ssel, Gian Maria Niccol? Benucci ?rta:

Hi Akos,

Thank you for the code. It worked. I reordered the dist objects as I 
cannot access the dataset of one of the two distance matrix.

dist_b_reordered <- as.dist(as.matrix(dist_b)[attr(dist_a, 

"Labels"), attr(dist_a, "Labels")])

all.equal(dist_a, dist_b_reordered, check.attributes = FALSE)

[1] "Mean relative difference: 0.8977224"

There is any statistical test that you suggest to compare the two 
distances? I would like to see if the two matrices show a similar 
pattern, e.g. mantel test?

Thanks,

Gian






On Sat, 1 Jun 2019 at 00:55, Bede-Fazekas ?kos <bfalevlist at gmail.com 
<mailto:bfalevlist at gmail.com>> wrote:

    Dear Gian,
    Let's do a small reproducible example:
    library(vegan)
    data(varespec)
    set.seed(12345)
    df_a <- t(varespec)
    df_b <- t(varespec[, sample.int
    <http://sample.int>(ncol(varespec))]) # shuffle the original
    column order
    dist_a <- vegdist(df_a); dist_b <- vegdist(df_b)
    all.equal(dist_a, dist_b, check.attributes = FALSE)

    Now, you have two options. If you have access to the original
    data.frame/matrix from which the dist object was created, you can
    reorder the rows of the second dataset to match that of the first
    one.
    Then do the distance matrix calculation again.
    df_b_reordered <- df_b[rownames(df_a), ]
    dist_b_reordered1 <- vegdist(df_b_reordered)
    all.equal(dist_a, dist_b_reordered1, check.attributes = FALSE)

    If you have no access to the original dataset and/or can not
    recalculate
    the distance matrix (e.g. due to computer capacity or time), you can
    reorder the distance matrix (both the rows and columns).
    dist_b_reordered2 <- as.dist(as.matrix(dist_b)[attr(dist_a,
    "Labels"),
    attr(dist_a, "Labels")])
    all.equal(dist_a, dist_b_reordered2, check.attributes = FALSE)

    HTH,
    ?kos Bede-Fazekas
    Hungarian Academy of Sciences


    2019.06.01. 1:53 keltez?ssel, Gian Maria Niccol? Benucci ?rta:

    > Hello,
    >
    > I have two vegan::vegdist objects I want to use into a

    vegan::mantel test.

    > However, rows and columns of the two objects are not in the same

    order. Is

    > there a way to reorder them one another?
    >
    > For example,
    >

    >> dist_a

    >? ? ? ? ? Lycovolu? Hupeaust? Lycofast? Lycodeut
    > Hupeaust 0.9045557
    >
    > Lycofast 0.6835441 0.8569781
    >
    > Lycodeut 0.8036577 0.8294657 0.7958988? ?...
    > ...
    >

    >> dist_b

    >? ? ? ? ? ?Lycodeut? ?Lycofast? ?Lycoscar? ?Lycovolu
    > Lycofast 0.09340659
    >
    > Lycoscar 0.08974359 0.01831502
    >
    > Lycovolu 0.09073476 0.01099655 0.01008080? ?...
    > ...
    >
    > Thanks,
    >

	[[alternative HTML version deleted]]

Hi Gian,
A Mantel test looks right for your problem. Check also protest in vegan.
Best,

Juan A. Balbuena

Dear Gian,
I'm not a statistician but software developer... hope that other members
of the mailing list are expert of distance matrices and can answer your
question.
Have a nice week,
?kos

2019.06.03. 17:44 keltez?ssel, Gian Maria Niccol? Benucci ?rta:

Hi Akos,

Thank you for the code. It worked. I reordered the dist objects as I
cannot access the dataset of one of the two distance matrix.

dist_b_reordered <- as.dist(as.matrix(dist_b)[attr(dist_a,

"Labels"), attr(dist_a, "Labels")])

all.equal(dist_a, dist_b_reordered, check.attributes = FALSE)

[1] "Mean relative difference: 0.8977224"

There is any statistical test that you suggest to compare the two
distances? I would like to see if the two matrices show a similar
pattern, e.g. mantel test?

Thanks,

Gian






On Sat, 1 Jun 2019 at 00:55, Bede-Fazekas ?kos <bfalevlist at gmail.com
<mailto:bfalevlist at gmail.com>> wrote:

    Dear Gian,
    Let's do a small reproducible example:
    library(vegan)
    data(varespec)
    set.seed(12345)
    df_a <- t(varespec)
    df_b <- t(varespec[, sample.int
    <http://sample.int>(ncol(varespec))]) # shuffle the original
    column order
    dist_a <- vegdist(df_a); dist_b <- vegdist(df_b)
    all.equal(dist_a, dist_b, check.attributes = FALSE)

    Now, you have two options. If you have access to the original
    data.frame/matrix from which the dist object was created, you can
    reorder the rows of the second dataset to match that of the first
    one.
    Then do the distance matrix calculation again.
    df_b_reordered <- df_b[rownames(df_a), ]
    dist_b_reordered1 <- vegdist(df_b_reordered)
    all.equal(dist_a, dist_b_reordered1, check.attributes = FALSE)

    If you have no access to the original dataset and/or can not
    recalculate
    the distance matrix (e.g. due to computer capacity or time), you

can

    reorder the distance matrix (both the rows and columns).
    dist_b_reordered2 <- as.dist(as.matrix(dist_b)[attr(dist_a,
    "Labels"),
    attr(dist_a, "Labels")])
    all.equal(dist_a, dist_b_reordered2, check.attributes = FALSE)

    HTH,
    ?kos Bede-Fazekas
    Hungarian Academy of Sciences


    2019.06.01. 1:53 keltez?ssel, Gian Maria Niccol? Benucci ?rta:

    > Hello,
    >
    > I have two vegan::vegdist objects I want to use into a

    vegan::mantel test.

    > However, rows and columns of the two objects are not in the same

    order. Is

    > there a way to reorder them one another?
    >
    > For example,
    >

    >> dist_a

    >          Lycovolu  Hupeaust  Lycofast  Lycodeut
    > Hupeaust 0.9045557
    >
    > Lycofast 0.6835441 0.8569781
    >
    > Lycodeut 0.8036577 0.8294657 0.7958988   ...
    > ...
    >

    >> dist_b

    >           Lycodeut   Lycofast   Lycoscar   Lycovolu
    > Lycofast 0.09340659
    >
    > Lycoscar 0.08974359 0.01831502
    >
    > Lycovolu 0.09073476 0.01099655 0.01008080   ...
    > ...
    >
    > Thanks,
    >

      [[alternative HTML version deleted]]

--
--
Dr. Juan A. Balbuena
Cavanilles Institute of Biodiversity and Evolutionary
Biology
University of Valencia
P.O. Box 22085
46071 Valencia, Spain
e-mail: j.a.balbuena at uv.es    tel. +34 963 543 658
+++++++++++++++++++++++++++++++++++++++++++++++++++
NOTE! For shipments by EXPRESS COURIER use the
following street address:
C/ Catedr?tico Jos? Beltr?n 2, 46980 Paterna
(Valencia), Spain.
+++++++++++++++++++++++++++++++++++++++++++++++++++