repr for flex arrays

Graeme.Winter＠Diamond.ac.uk

14 Sep 2018 14 Sep '18

2:29 p.m.

Hi Folks, Is there a reason for not having a helpful __repr__ for flex arrays i.e. Grey-Area:002-decide-when graeme$ dials.python Python 2.7.14 (default, Apr 17 2018, 15:02:10) [GCC 4.2.1 Compatible Apple LLVM 9.1.0 (clang-902.0.39.1)] on darwin Type "help", "copyright", "credits" or "license" for more information.

...

...
...
from dials.array_family import flex a = flex.double([1, 2, 3]) print a

(I have a feeling I asked this before but cannot find reference) For example

...

...
...
print a.as_numpy_array() [ 1. 2. 3.]

is more useful… but annoying to type I appreciate a could be “big” but numpy seems to handle this OK even - could __repr__ be punned to as_numpy_array(),__repr__()? Currently it prints the type of the array and its location in ram, which are adequately served by

...

...
...
type(a) print id(a) 4451627968

If I print an array I more often want to see the content and less often the storage class and location (I humbly assert) Thoughts? I know this tends to confuse and annoy people who are new to flex arrays (and veterans who are busy debugging stuff…) Many thanks Graeme -- This e-mail and any attachments may contain confidential, copyright and or privileged material, and are for the use of the intended addressee only. If you are not the intended addressee or an authorised recipient of the addressee please notify us of receipt by returning the e-mail and do not use, copy, retain, distribute or disclose the information in or attached to the e-mail. Any opinions expressed within this e-mail are those of the individual and not necessarily of Diamond Light Source Ltd. Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments are free from viruses and we cannot accept liability for any damage which you may sustain as a result of software viruses which may be transmitted in or with the message. Diamond Light Source Limited (company no. 4375679). Registered in England and Wales with its registered office at Diamond House, Harwell Science and Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom

Show replies by date

Aaron Brewster

14 Sep 14 Sep

4:23 p.m.

Hi Graeme, I'm not sure what the right answer here is except I don't think it is as_numpy_array().__repr__ as that does a copy. No good for big arrays. On Fri, Sep 14, 2018 at 7:30 AM [email protected] < [email protected]> wrote:

...

Hi Folks,

Is there a reason for not having a helpful __repr__ for flex arrays i.e.

Grey-Area:002-decide-when graeme$ dials.python Python 2.7.14 (default, Apr 17 2018, 15:02:10) [GCC 4.2.1 Compatible Apple LLVM 9.1.0 (clang-902.0.39.1)] on darwin Type "help", "copyright", "credits" or "license" for more information.

...
...
...
from dials.array_family import flex a = flex.double([1, 2, 3]) print a

(I have a feeling I asked this before but cannot find reference)

For example

...
...
...
print a.as_numpy_array() [ 1. 2. 3.]

is more useful… but annoying to type

I appreciate a could be “big” but numpy seems to handle this OK

even - could __repr__ be punned to as_numpy_array(),__repr__()?

Currently it prints the type of the array and its location in ram, which are adequately served by

...
...
...
type(a) print id(a) 4451627968

If I print an array I more often want to see the content and less often the storage class and location (I humbly assert)

Thoughts? I know this tends to confuse and annoy people who are new to flex arrays (and veterans who are busy debugging stuff…)

Many thanks Graeme

-- This e-mail and any attachments may contain confidential, copyright and or privileged material, and are for the use of the intended addressee only. If you are not the intended addressee or an authorised recipient of the addressee please notify us of receipt by returning the e-mail and do not use, copy, retain, distribute or disclose the information in or attached to the e-mail. Any opinions expressed within this e-mail are those of the individual and not necessarily of Diamond Light Source Ltd. Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments are free from viruses and we cannot accept liability for any damage which you may sustain as a result of software viruses which may be transmitted in or with the message. Diamond Light Source Limited (company no. 4375679). Registered in England and Wales with its registered office at Diamond House, Harwell Science and Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

Graeme.Winter＠Diamond.ac.uk

4:27 p.m.

Does it? Dang! Thought it would be fly weight and just make the data structures needed. If that’s the case then completely agree. Cheers Graeme ________________________________ From: [email protected] on behalf of Aaron Brewster Sent: 14 September 2018 17:23:41 To: cctbx mailing list Subject: Re: [cctbxbb] repr for flex arrays Hi Graeme, I'm not sure what the right answer here is except I don't think it is as_numpy_array().__repr__ as that does a copy. No good for big arrays. On Fri, Sep 14, 2018 at 7:30 AM [email protected]mailto:[email protected] mailto:[email protected]> wrote: Hi Folks, Is there a reason for not having a helpful __repr__ for flex arrays i.e. Grey-Area:002-decide-when graeme$ dials.python Python 2.7.14 (default, Apr 17 2018, 15:02:10) [GCC 4.2.1 Compatible Apple LLVM 9.1.0 (clang-902.0.39.1)] on darwin Type "help", "copyright", "credits" or "license" for more information.

...

...
...
from dials.array_family import flex a = flex.double([1, 2, 3]) print a

(I have a feeling I asked this before but cannot find reference) For example

...

...
...
print a.as_numpy_array() [ 1. 2. 3.]

...

...
...
type(a) print id(a) 4451627968

Nicholas Devenish

4:56 p.m.

I think as_numpy_array was just a suggestion of the simplest possible starting implementation. But is there a philosophical objection to them being added? Nick On Fri, Sep 14, 2018 at 5:27 PM [email protected] wrote:

...

Does it? Dang! Thought it would be fly weight and just make the data structures needed.

If that’s the case then completely agree.

Cheers Graeme ________________________________ From: [email protected] on behalf of Aaron Brewster Sent: 14 September 2018 17:23:41 To: cctbx mailing list Subject: Re: [cctbxbb] repr for flex arrays

Hi Graeme, I'm not sure what the right answer here is except I don't think it is as_numpy_array().__repr__ as that does a copy. No good for big arrays.

On Fri, Sep 14, 2018 at 7:30 AM [email protected]mailto:[email protected] mailto:[email protected]> wrote: Hi Folks,

Is there a reason for not having a helpful __repr__ for flex arrays i.e.

Grey-Area:002-decide-when graeme$ dials.python Python 2.7.14 (default, Apr 17 2018, 15:02:10) [GCC 4.2.1 Compatible Apple LLVM 9.1.0 (clang-902.0.39.1)] on darwin Type "help", "copyright", "credits" or "license" for more information.

...
...
...
from dials.array_family import flex a = flex.double([1, 2, 3]) print a

(I have a feeling I asked this before but cannot find reference)

For example

...
...
...
print a.as_numpy_array() [ 1. 2. 3.]

is more useful… but annoying to type

I appreciate a could be “big” but numpy seems to handle this OK

even - could __repr__ be punned to as_numpy_array(),__repr__()?

Currently it prints the type of the array and its location in ram, which are adequately served by

...
...
...
type(a) print id(a) 4451627968

If I print an array I more often want to see the content and less often the storage class and location (I humbly assert)

Thoughts? I know this tends to confuse and annoy people who are new to flex arrays (and veterans who are busy debugging stuff…)

Many thanks Graeme

-- This e-mail and any attachments may contain confidential, copyright and or privileged material, and are for the use of the intended addressee only. If you are not the intended addressee or an authorised recipient of the addressee please notify us of receipt by returning the e-mail and do not use, copy, retain, distribute or disclose the information in or attached to the e-mail. Any opinions expressed within this e-mail are those of the individual and not necessarily of Diamond Light Source Ltd. Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments are free from viruses and we cannot accept liability for any damage which you may sustain as a result of software viruses which may be transmitted in or with the message. Diamond Light Source Limited (company no. 4375679). Registered in England and Wales with its registered office at Diamond House, Harwell Science and Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom

_______________________________________________ cctbxbb mailing list [email protected]mailto:[email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

Graeme.Winter＠Diamond.ac.uk

5:08 p.m.

Yes, my numpy suggestion was probably oversimplified, but the question remains whether there is a reason for the current behaviour Thanks Graeme ________________________________ From: [email protected] on behalf of Nicholas Devenish Sent: 14 September 2018 17:56:26 To: cctbx mailing list Subject: Re: [cctbxbb] repr for flex arrays I think as_numpy_array was just a suggestion of the simplest possible starting implementation. But is there a philosophical objection to them being added? Nick On Fri, Sep 14, 2018 at 5:27 PM [email protected] wrote:

...

Does it? Dang! Thought it would be fly weight and just make the data structures needed.

If that’s the case then completely agree.

Cheers Graeme ________________________________ From: [email protected] on behalf of Aaron Brewster Sent: 14 September 2018 17:23:41 To: cctbx mailing list Subject: Re: [cctbxbb] repr for flex arrays

Hi Graeme, I'm not sure what the right answer here is except I don't think it is as_numpy_array().__repr__ as that does a copy. No good for big arrays.

On Fri, Sep 14, 2018 at 7:30 AM [email protected]mailto:[email protected] mailto:[email protected]> wrote: Hi Folks,

Is there a reason for not having a helpful __repr__ for flex arrays i.e.

Grey-Area:002-decide-when graeme$ dials.python Python 2.7.14 (default, Apr 17 2018, 15:02:10) [GCC 4.2.1 Compatible Apple LLVM 9.1.0 (clang-902.0.39.1)] on darwin Type "help", "copyright", "credits" or "license" for more information.

...
...
...
from dials.array_family import flex a = flex.double([1, 2, 3]) print a

(I have a feeling I asked this before but cannot find reference)

For example

...
...
...
print a.as_numpy_array() [ 1. 2. 3.]

is more useful… but annoying to type

I appreciate a could be “big” but numpy seems to handle this OK

even - could __repr__ be punned to as_numpy_array(),__repr__()?

Currently it prints the type of the array and its location in ram, which are adequately served by

...
...
...
type(a) print id(a) 4451627968

If I print an array I more often want to see the content and less often the storage class and location (I humbly assert)

Thoughts? I know this tends to confuse and annoy people who are new to flex arrays (and veterans who are busy debugging stuff…)

Many thanks Graeme

-- This e-mail and any attachments may contain confidential, copyright and or privileged material, and are for the use of the intended addressee only. If you are not the intended addressee or an authorised recipient of the addressee please notify us of receipt by returning the e-mail and do not use, copy, retain, distribute or disclose the information in or attached to the e-mail. Any opinions expressed within this e-mail are those of the individual and not necessarily of Diamond Light Source Ltd. Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments are free from viruses and we cannot accept liability for any damage which you may sustain as a result of software viruses which may be transmitted in or with the message. Diamond Light Source Limited (company no. 4375679). Registered in England and Wales with its registered office at Diamond House, Harwell Science and Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom

_______________________________________________ cctbxbb mailing list [email protected]mailto:[email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

Nicholas Sauter

5:49 p.m.

Yes, there is a philosophical objection to using .as_numpy_array(). Numpy only deals with a fixed number of data types, such as int, double, etc. Flex arrays can be extended to many other types such as Bragg spots, 2-vectors, 3-vectors, and 3x3 matrices. Therefore, if we are going to take the performance hit and do a copy, I would simply recommend using a Python list: __repr__(a) :== list(a) Nick Nicholas K. Sauter, Ph. D. Senior Scientist, Molecular Biophysics & Integrated Bioimaging Division Lawrence Berkeley National Laboratory 1 Cyclotron Rd., Bldg. 33R0345 Berkeley, CA 94720 (510) 486-5713 On Fri, Sep 14, 2018 at 9:56 AM, Nicholas Devenish wrote:

...

I think as_numpy_array was just a suggestion of the simplest possible starting implementation.

But is there a philosophical objection to them being added?

Nick

On Fri, Sep 14, 2018 at 5:27 PM [email protected] wrote:

...
Does it? Dang! Thought it would be fly weight and just make the data

structures needed.

...
If that’s the case then completely agree.

Cheers Graeme ________________________________ From: [email protected]
...
Sent: 14 September 2018 17:23:41 To: cctbx mailing list Subject: Re: [cctbxbb] repr for flex arrays

Hi Graeme, I'm not sure what the right answer here is except I don't

online.org> on behalf of Aaron Brewster think it is as_numpy_array().__repr__ as that does a copy. No good for big arrays.

...
On Fri, Sep 14, 2018 at 7:30 AM [email protected]
...
Hi Folks,

Is there a reason for not having a helpful __repr__ for flex arrays i.e.

Grey-Area:002-decide-when graeme$ dials.python Python 2.7.14 (default, Apr 17 2018, 15:02:10) [GCC 4.2.1 Compatible Apple LLVM 9.1.0 (clang-902.0.39.1)] on darwin Type "help", "copyright", "credits" or "license" for more information.

...
...
...
from dials.array_family import flex a = flex.double([1, 2, 3]) print a

(I have a feeling I asked this before but cannot find reference)

For example

...
...
...
print a.as_numpy_array() [ 1. 2. 3.]

is more useful… but annoying to type

I appreciate a could be “big” but numpy seems to handle this OK

even - could __repr__ be punned to as_numpy_array(),__repr__()?

Currently it prints the type of the array and its location in ram, which are adequately served by

...
...
...
type(a) print id(a) 4451627968

If I print an array I more often want to see the content and less often

[email protected]> > wrote: the storage class and location (I humbly assert)

...
Thoughts? I know this tends to confuse and annoy people who are new to

flex arrays (and veterans who are busy debugging stuff…)

...
Many thanks Graeme

-- This e-mail and any attachments may contain confidential, copyright and

...
Any opinions expressed within this e-mail are those of the individual and not necessarily of Diamond Light Source Ltd. Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments are free from viruses and we cannot accept liability for any damage which you may sustain as a result of software viruses which may be

or privileged material, and are for the use of the intended addressee only. If you are not the intended addressee or an authorised recipient of the addressee please notify us of receipt by returning the e-mail and do not use, copy, retain, distribute or disclose the information in or attached to the e-mail. transmitted in or with the message.

...
Diamond Light Source Limited (company no. 4375679). Registered in England and Wales with its registered office at Diamond House, Harwell Science and Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom

_______________________________________________ cctbxbb mailing list [email protected]mailto:[email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

Aaron Brewster

6 p.m.

Hi, I think avoiding a copy is good. What about this: Flex type , focus: , size: memory address <address>, first and last element if len(a)>0: , Like numpy does, let's avoid printing every element. On Fri, Sep 14, 2018 at 10:50 AM Nicholas Sauter wrote:

...

Yes, there is a philosophical objection to using .as_numpy_array(). Numpy only deals with a fixed number of data types, such as int, double, etc. Flex arrays can be extended to many other types such as Bragg spots, 2-vectors, 3-vectors, and 3x3 matrices. Therefore, if we are going to take the performance hit and do a copy, I would simply recommend using a Python list:

__repr__(a) :== list(a)

Nick

Nicholas K. Sauter, Ph. D. Senior Scientist, Molecular Biophysics & Integrated Bioimaging Division Lawrence Berkeley National Laboratory 1 Cyclotron Rd., Bldg. 33R0345 Berkeley, CA 94720 (510) 486-5713

On Fri, Sep 14, 2018 at 9:56 AM, Nicholas Devenish wrote:

...
I think as_numpy_array was just a suggestion of the simplest possible starting implementation.

But is there a philosophical objection to them being added?

Nick

On Fri, Sep 14, 2018 at 5:27 PM [email protected] wrote:

...
Does it? Dang! Thought it would be fly weight and just make the data

structures needed.

...
If that’s the case then completely agree.

Cheers Graeme ________________________________ From: [email protected] <

...
Sent: 14 September 2018 17:23:41 To: cctbx mailing list Subject: Re: [cctbxbb] repr for flex arrays

Hi Graeme, I'm not sure what the right answer here is except I don't

[email protected]> on behalf of Aaron Brewster < [email protected]> think it is as_numpy_array().__repr__ as that does a copy. No good for big arrays.

...
On Fri, Sep 14, 2018 at 7:30 AM [email protected]
...
Hi Folks,

Is there a reason for not having a helpful __repr__ for flex arrays i.e.

Grey-Area:002-decide-when graeme$ dials.python Python 2.7.14 (default, Apr 17 2018, 15:02:10) [GCC 4.2.1 Compatible Apple LLVM 9.1.0 (clang-902.0.39.1)] on darwin Type "help", "copyright", "credits" or "license" for more information.

...
...
...
from dials.array_family import flex a = flex.double([1, 2, 3]) print a

(I have a feeling I asked this before but cannot find reference)

For example

...
...
...
print a.as_numpy_array() [ 1. 2. 3.]

is more useful… but annoying to type

I appreciate a could be “big” but numpy seems to handle this OK

even - could __repr__ be punned to as_numpy_array(),__repr__()?

Currently it prints the type of the array and its location in ram, which are adequately served by

...
...
...
type(a) print id(a) 4451627968

If I print an array I more often want to see the content and less often

[email protected]> > wrote: the storage class and location (I humbly assert)

...
Thoughts? I know this tends to confuse and annoy people who are new to

flex arrays (and veterans who are busy debugging stuff…)

...
Many thanks Graeme

-- This e-mail and any attachments may contain confidential, copyright and

...
Any opinions expressed within this e-mail are those of the individual and not necessarily of Diamond Light Source Ltd. Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments are free from viruses and we cannot accept liability for any damage which you may sustain as a result of software viruses which may be

or privileged material, and are for the use of the intended addressee only. If you are not the intended addressee or an authorised recipient of the addressee please notify us of receipt by returning the e-mail and do not use, copy, retain, distribute or disclose the information in or attached to the e-mail. transmitted in or with the message.

...
Diamond Light Source Limited (company no. 4375679). Registered in England and Wales with its registered office at Diamond House, Harwell Science and Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom

_______________________________________________ cctbxbb mailing list [email protected]mailto:[email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

Graeme.Winter＠Diamond.ac.uk

6:08 p.m.

Hi Aaron Yes I think this would be useful. I’d perhaps be more permissive say printing 10 or fewer items, but this is is the direction The numpy illustration I gave was an unfortunate choice. Thanks Graeme On 14 Sep 2018, at 19:01, Aaron Brewster mailto:[email protected]> wrote: Hi, I think avoiding a copy is good. What about this: Flex type , focus: , size: memory address <address>, first and last element if len(a)>0: , Like numpy does, let's avoid printing every element. On Fri, Sep 14, 2018 at 10:50 AM Nicholas Sauter mailto:[email protected]> wrote: Yes, there is a philosophical objection to using .as_numpy_array(). Numpy only deals with a fixed number of data types, such as int, double, etc. Flex arrays can be extended to many other types such as Bragg spots, 2-vectors, 3-vectors, and 3x3 matrices. Therefore, if we are going to take the performance hit and do a copy, I would simply recommend using a Python list: __repr__(a) :== list(a) Nick Nicholas K. Sauter, Ph. D. Senior Scientist, Molecular Biophysics & Integrated Bioimaging Division Lawrence Berkeley National Laboratory 1 Cyclotron Rd., Bldg. 33R0345 Berkeley, CA 94720 (510) 486-5713 On Fri, Sep 14, 2018 at 9:56 AM, Nicholas Devenish mailto:[email protected]> wrote: I think as_numpy_array was just a suggestion of the simplest possible starting implementation. But is there a philosophical objection to them being added? Nick On Fri, Sep 14, 2018 at 5:27 PM [email protected]mailto:[email protected] mailto:[email protected]> wrote:

...

Does it? Dang! Thought it would be fly weight and just make the data structures needed.

If that’s the case then completely agree.

Cheers Graeme ________________________________ From: [email protected]mailto:[email protected] mailto:[email protected]> on behalf of Aaron Brewster mailto:[email protected]> Sent: 14 September 2018 17:23:41 To: cctbx mailing list Subject: Re: [cctbxbb] repr for flex arrays

Hi Graeme, I'm not sure what the right answer here is except I don't think it is as_numpy_array().__repr__ as that does a copy. No good for big arrays.

On Fri, Sep 14, 2018 at 7:30 AM [email protected]mailto:[email protected]mailto:[email protected]> mailto:[email protected]mailto:[email protected]>> wrote: Hi Folks,

Is there a reason for not having a helpful __repr__ for flex arrays i.e.

Grey-Area:002-decide-when graeme$ dials.python Python 2.7.14 (default, Apr 17 2018, 15:02:10) [GCC 4.2.1 Compatible Apple LLVM 9.1.0 (clang-902.0.39.1)] on darwin Type "help", "copyright", "credits" or "license" for more information.

...
...
...
from dials.array_family import flex a = flex.double([1, 2, 3]) print a

(I have a feeling I asked this before but cannot find reference)

For example

...
...
...
print a.as_numpy_array() [ 1. 2. 3.]

is more useful… but annoying to type

I appreciate a could be “big” but numpy seems to handle this OK

even - could __repr__ be punned to as_numpy_array(),__repr__()?

Currently it prints the type of the array and its location in ram, which are adequately served by

...
...
...
type(a) print id(a) 4451627968

If I print an array I more often want to see the content and less often the storage class and location (I humbly assert)

Thoughts? I know this tends to confuse and annoy people who are new to flex arrays (and veterans who are busy debugging stuff…)

Many thanks Graeme

-- This e-mail and any attachments may contain confidential, copyright and or privileged material, and are for the use of the intended addressee only. If you are not the intended addressee or an authorised recipient of the addressee please notify us of receipt by returning the e-mail and do not use, copy, retain, distribute or disclose the information in or attached to the e-mail. Any opinions expressed within this e-mail are those of the individual and not necessarily of Diamond Light Source Ltd. Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments are free from viruses and we cannot accept liability for any damage which you may sustain as a result of software viruses which may be transmitted in or with the message. Diamond Light Source Limited (company no. 4375679). Registered in England and Wales with its registered office at Diamond House, Harwell Science and Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom

_______________________________________________ cctbxbb mailing list [email protected]mailto:[email protected]mailto:[email protected]> http://phenix-online.org/mailman/listinfo/cctbxbb

_______________________________________________ cctbxbb mailing list [email protected]mailto:[email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

_______________________________________________ cctbxbb mailing list [email protected]mailto:[email protected] http://phenix-online.org/mailman/listinfo/cctbxbb _______________________________________________ cctbxbb mailing list [email protected]mailto:[email protected] http://phenix-online.org/mailman/listinfo/cctbxbb _______________________________________________ cctbxbb mailing list [email protected]mailto:[email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

Nicholas Devenish

6:09 p.m.

I wasn't asking if there was a philosophical objection to as_numpy_array. I was asking if there was a philosophical objection to having a __repr__ on flex arrays. This is because I seem to remember finding instances where they had been actively removed before. Nick On Fri, Sep 14, 2018 at 7:01 PM Aaron Brewster wrote:

...

Hi, I think avoiding a copy is good. What about this:

Flex type , focus: , size: memory address <address>, first and last element if len(a)>0: ,

Like numpy does, let's avoid printing every element.

On Fri, Sep 14, 2018 at 10:50 AM Nicholas Sauter wrote:

...
Yes, there is a philosophical objection to using .as_numpy_array(). Numpy only deals with a fixed number of data types, such as int, double, etc. Flex arrays can be extended to many other types such as Bragg spots, 2-vectors, 3-vectors, and 3x3 matrices. Therefore, if we are going to take the performance hit and do a copy, I would simply recommend using a Python list:

__repr__(a) :== list(a)

Nick

Nicholas K. Sauter, Ph. D. Senior Scientist, Molecular Biophysics & Integrated Bioimaging Division Lawrence Berkeley National Laboratory 1 Cyclotron Rd., Bldg. 33R0345 Berkeley, CA 94720 (510) 486-5713

On Fri, Sep 14, 2018 at 9:56 AM, Nicholas Devenish wrote:

...
I think as_numpy_array was just a suggestion of the simplest possible starting implementation.

But is there a philosophical objection to them being added?

Nick

On Fri, Sep 14, 2018 at 5:27 PM [email protected] wrote:

...
Does it? Dang! Thought it would be fly weight and just make the data structures needed.

If that’s the case then completely agree.

Cheers Graeme ________________________________ From: [email protected] on behalf of Aaron Brewster Sent: 14 September 2018 17:23:41 To: cctbx mailing list Subject: Re: [cctbxbb] repr for flex arrays

Hi Graeme, I'm not sure what the right answer here is except I don't think it is as_numpy_array().__repr__ as that does a copy. No good for big arrays.

On Fri, Sep 14, 2018 at 7:30 AM [email protected]mailto:[email protected] mailto:[email protected]> wrote: Hi Folks,

Is there a reason for not having a helpful __repr__ for flex arrays i.e.

Grey-Area:002-decide-when graeme$ dials.python Python 2.7.14 (default, Apr 17 2018, 15:02:10) [GCC 4.2.1 Compatible Apple LLVM 9.1.0 (clang-902.0.39.1)] on darwin Type "help", "copyright", "credits" or "license" for more information.

...
...
> from dials.array_family import flex > a = flex.double([1, 2, 3]) > print a

(I have a feeling I asked this before but cannot find reference)

For example

...
...
> print a.as_numpy_array() [ 1. 2. 3.]

is more useful… but annoying to type

I appreciate a could be “big” but numpy seems to handle this OK

even - could __repr__ be punned to as_numpy_array(),__repr__()?

Currently it prints the type of the array and its location in ram, which are adequately served by

...
...
> type(a) > print id(a) 4451627968

If I print an array I more often want to see the content and less often the storage class and location (I humbly assert)

Thoughts? I know this tends to confuse and annoy people who are new to flex arrays (and veterans who are busy debugging stuff…)

Many thanks Graeme

-- This e-mail and any attachments may contain confidential, copyright and or privileged material, and are for the use of the intended addressee only. If you are not the intended addressee or an authorised recipient of the addressee please notify us of receipt by returning the e-mail and do not use, copy, retain, distribute or disclose the information in or attached to the e-mail. Any opinions expressed within this e-mail are those of the individual and not necessarily of Diamond Light Source Ltd. Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments are free from viruses and we cannot accept liability for any damage which you may sustain as a result of software viruses which may be transmitted in or with the message. Diamond Light Source Limited (company no. 4375679). Registered in England and Wales with its registered office at Diamond House, Harwell Science and Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom

_______________________________________________ cctbxbb mailing list [email protected]mailto:[email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

Graeme.Winter＠Diamond.ac.uk

6:11 p.m.

Nick This was a simple illustration of the behaviour rather than a real implementation suggestion. But as Aaron suggested we could make the representation more helpful Anything to avoid a copy to a list or numpy array Cheers Graeme On 14 Sep 2018, at 18:50, Nicholas Sauter mailto:[email protected]> wrote: Yes, there is a philosophical objection to using .as_numpy_array(). Numpy only deals with a fixed number of data types, such as int, double, etc. Flex arrays can be extended to many other types such as Bragg spots, 2-vectors, 3-vectors, and 3x3 matrices. Therefore, if we are going to take the performance hit and do a copy, I would simply recommend using a Python list: __repr__(a) :== list(a) Nick Nicholas K. Sauter, Ph. D. Senior Scientist, Molecular Biophysics & Integrated Bioimaging Division Lawrence Berkeley National Laboratory 1 Cyclotron Rd., Bldg. 33R0345 Berkeley, CA 94720 (510) 486-5713 On Fri, Sep 14, 2018 at 9:56 AM, Nicholas Devenish mailto:[email protected]> wrote: I think as_numpy_array was just a suggestion of the simplest possible starting implementation. But is there a philosophical objection to them being added? Nick On Fri, Sep 14, 2018 at 5:27 PM [email protected]mailto:[email protected] mailto:[email protected]> wrote:

...

Does it? Dang! Thought it would be fly weight and just make the data structures needed.

If that’s the case then completely agree.

Cheers Graeme ________________________________ From: [email protected]mailto:[email protected] mailto:[email protected]> on behalf of Aaron Brewster mailto:[email protected]> Sent: 14 September 2018 17:23:41 To: cctbx mailing list Subject: Re: [cctbxbb] repr for flex arrays

Hi Graeme, I'm not sure what the right answer here is except I don't think it is as_numpy_array().__repr__ as that does a copy. No good for big arrays.

On Fri, Sep 14, 2018 at 7:30 AM [email protected]mailto:[email protected]mailto:[email protected]> mailto:[email protected]mailto:[email protected]>> wrote: Hi Folks,

Is there a reason for not having a helpful __repr__ for flex arrays i.e.

Grey-Area:002-decide-when graeme$ dials.python Python 2.7.14 (default, Apr 17 2018, 15:02:10) [GCC 4.2.1 Compatible Apple LLVM 9.1.0 (clang-902.0.39.1)] on darwin Type "help", "copyright", "credits" or "license" for more information.

...
...
...
from dials.array_family import flex a = flex.double([1, 2, 3]) print a

(I have a feeling I asked this before but cannot find reference)

For example

...
...
...
print a.as_numpy_array() [ 1. 2. 3.]

is more useful… but annoying to type

I appreciate a could be “big” but numpy seems to handle this OK

even - could __repr__ be punned to as_numpy_array(),__repr__()?

Currently it prints the type of the array and its location in ram, which are adequately served by

...
...
...
type(a) print id(a) 4451627968

If I print an array I more often want to see the content and less often the storage class and location (I humbly assert)

Thoughts? I know this tends to confuse and annoy people who are new to flex arrays (and veterans who are busy debugging stuff…)

Many thanks Graeme

-- This e-mail and any attachments may contain confidential, copyright and or privileged material, and are for the use of the intended addressee only. If you are not the intended addressee or an authorised recipient of the addressee please notify us of receipt by returning the e-mail and do not use, copy, retain, distribute or disclose the information in or attached to the e-mail. Any opinions expressed within this e-mail are those of the individual and not necessarily of Diamond Light Source Ltd. Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments are free from viruses and we cannot accept liability for any damage which you may sustain as a result of software viruses which may be transmitted in or with the message. Diamond Light Source Limited (company no. 4375679). Registered in England and Wales with its registered office at Diamond House, Harwell Science and Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom

_______________________________________________ cctbxbb mailing list [email protected]mailto:[email protected]mailto:[email protected]> http://phenix-online.org/mailman/listinfo/cctbxbb

_______________________________________________ cctbxbb mailing list [email protected]mailto:[email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

2132

Age (days ago)

2132

Last active (days ago)

List overview

Download

9 comments

4 participants

participants (4)

Aaron Brewster
Graeme.Winter＠Diamond.ac.uk
Nicholas Devenish
Nicholas Sauter

repr for flex arrays

tags

participants (4)