Re: GSOC2018 - Rubyplot 2018-6-15 Working on Bar Charts

Arafat Khan

unread,

Jun 18, 2018, 2:19:07 AM6/18/18

to Prasun Anand, John Woods, Sameer Deshmukh, Pjotr Prins, Pranav Garg, SciRuby Mailing List

Okay then, I will probably write them in Github pages and then at the end I will write a compilation in the main blog post.

Content is certainly not an issue for blog post.

My concern is the level of interest generated by a general audience in a blog post.

So Basically sharing a well-developed software via blog post on reddit gets a lot of traffic.

It won't be the case when you do things in other ways.

Regards,

Arafat

On Sun, Jun 17, 2018 at 11:15 PM Prasun Anand <prasunan...@gmail.com> wrote:

Arafat,

You need to write at least 2 blog posts in a month because I believe that you have a story to tell regarding
the work you have done. 80 hours of work in 2 weeks should easily result in a great post.

Why do you need to write regular posts ?
https://groups.google.com/forum/#!topic/sciruby-dev/x3xlGKcCCK4

To add to this, this helps other students and researchers get involved with the project.

Regards,
Prasun

On Mon, Jun 18, 2018 at 6:24 AM Arafat Khan <arafat....@gmail.com> wrote:

On Sun, Jun 17, 2018 at 5:29 PM Arafat Khan <arafat....@gmail.com> wrote:
Hi Prasun,
I have mentioned before in my emails, that I don't have any issues writing a blog post.
I just don't want to write a blog post for the sake of it. I am happy to document my progress and discuss it among ourselves in elaborate emails.
Usually, I use my blog to introduce software and write interesting ideas to build similar things( https://medium.com/@Arafat./ ). I would very much like to
write 2-3 blog posts after I have made some reasonable progress(i.e. possibly after completing the Magick plot and integrating GR with it).

Regards,
Arafat

On Sun, Jun 10, 2018 at 11:13 AM Arafat Khan <arafat....@gmail.com> wrote:
Hi Everyone,
I am really sorry for the super delayed update. I have been super sick
this week. I haven't been able to do much work in the past few days as
I had a high fever. I have however tried my best to move forward. I added the
API to make scatter plots and that looks pretty cool. I will have to test it
a bit more but it's still some good plotting idea and I think that I will be able to easily
add a lot more plotting types as I am getting the gist of all these things very well
so building up on my previous ideas will be super easy and fun.

Overall, I would say that the work till now has been good and
I have been able to add quite a few useful functionalities as well as being
able to stress test a few of the past work. I have a much clearer idea of what I need to do
for the future of this project and how I will go about it.

As for the blog posts, I would say that I have written a few blog posts before for introducing my
software and to communicate with Ruby and Golang community. The reason
I am not writing a blog post right now is because if I wanted to communicate about my work
then I can very well do that by emailing you guys and other developers involved
with plotting and visualizations.

In future, I will be writing 2-3 very elaborate blog posts and share
them with the community later on Reddit or hackernews and see how people react
to be able to take the feedback to add more features. I will do that later once
I have had some reasonable progress.

Please do let me know if you need to know anything else or have any questions
about the work.
Regards,
Arafat

On Tue, Jun 5, 2018 at 10:55 AM Arafat Khan <arafat....@gmail.com> wrote:
Yes, that is the reason for me not implementing the transformations.
Personally, I am just doing simple operations on the data first to be able to make simple plots first.
If this works out well then I could consider adding more features(transformations).

My code is very modular so adding that shouldn't be an issue.

Regards,
Arafat

On Sun, Jun 17, 2018 at 3:47 AM Prasun Anand <prasunan...@gmail.com> wrote:
Hi Arafat,

I don't see any blog posts from you.

When can we expect to see it ?

Regards,
Prasun

On Sat, Jun 16, 2018, 2:36 AM Arafat Khan <arafat....@gmail.com> wrote:
Hi Everyone,
I read the first evaluation and I have been rather disappointed but then again it's kinda my fault for the lack of communication so I am really sorry about this.

I can only say that I have been spending 9+ hours every weekday on the project(with a few exceptions of health issues) and working extra hard because it's really interesting way to learn new things.

Very early on in the project, I just ended up hitting a stage where I really can't make much use of external code and things learned from sources because the Matplotlib code base is too evolved to be easily replicated. I have resort things the good old-fashioned way of making things from scratch and literally, make the 2-d drawings by just hardcoding and then generalizing things based on data.

There have been many interesting questions for every graph type, take for example bar graph
1. How do you make sure that a super large bar data point just doesn't go "off" the image plane?
2. How much spacing based on the number of different bars to be made?
3. How to do labeling and where to save space in the image for it?
4. How make sure the each and every object doesn't overlap or get into another's space?

This has been pretty hard and involves a lot of math instead of any other technical problem. I didn't bother emailing you for most of the days because I really didn't have much to show as a good result for most days. Anyways, I have decided to email you guys every weekday(Monday-Friday) no matter how things go just so that we are all updated on the work. I hope this is okay with you guys.

As for the screenshot that I sent looks perfectly okay and currently drawing the
bars itself ain't the problem it's the normalization of data that troubled me.
Also, another issue currently was that the bar graph that I sent you only had
all the bars going in the positive direction so that's not cool. I am adding methods to make the additions into both positive and negative directions. I will try my best to commit the whole thing by tomorrow.

Regards,
Arafat

On Thu, Jun 14, 2018 at 10:23 PM Arafat Khan <arafat....@gmail.com> wrote:
Sure, here it is.
The colors are kinda messed up right now but that can be taken care of..

On Thu, Jun 14, 2018 at 12:13 PM John Woods <john.o...@gmail.com> wrote:
Want to send us a screenshot?

On Thu, Jun 14, 2018 at 10:47 AM Arafat Khan <arafat....@gmail.com> wrote:
Hi Everyone,
I am working on bar charts right now and things are looking better than yesterday.
I am able to draw the rectangular bars and still working on adjusting the spacing in the bars to make things work well. I hope I am able to finish this by the end of the week. Currently, I haven't focussed on the data normalization aspect but that would certainly influence the outlook of the graph. So first, I am just looking at making the rectangle bars and focusing on spacing etc on some test data.

Regards,
Arafat

On Wed, Jun 13, 2018 at 8:52 PM John Woods <john.o...@gmail.com> wrote:
I'd say don't worry too much about overlapping labels (at least not until much, much later). I think avoiding collisions is a whole GSOC project on its own.

On Tue, Jun 12, 2018 at 12:09 PM Arafat Khan <arafat....@gmail.com> wrote:
Hi Everyone,
I am sorry but I haven't been updating you guys daily. I will take care of that for now on and update you guys on every weekday. Healthwise, I am doing much better now.

For the past few days, I have just been thinking of the math behind alignment. I guess the real problem for me is having the corner cases taken care of. I think that at the very least everything that involves the base/vertical corner lines is okay. So the axis and vertical and horizontal bars are fine. The X-Axis, Y-Axis Title naming is going to be fine too.

There may, however, be a few problems when someone tries to set up too many labels for data. I just don't want those weird looking images that have everything overlapping because it looks very unaesthetic. Perhaps I will just set up some error conditions to warn users in those cases.

Another major concern is the proper testing mechanism. I never bothered focussing on this because it doesn't make much of a difference given the stage of development and considering that I am yet to make many changes to the plots. I will be reading a bit of https://github.com/matplotlib/pytest-mpl.
Right now this week, I am mainly focussing on making bar charts.

Regards,
Arafat

Prasun Anand

unread,

Jun 18, 2018, 2:53:33 AM6/18/18

to Arafat Khan, John Woods, Sameer Deshmukh, Pjotr Prins, Pranav Garg, SciRuby Mailing List

This is not Tensorflow :)

You will not get enough traffic from Reddit.

Sameer, Please clarify.

Regards,

Prasun

Arafat Khan

unread,

Jun 18, 2018, 2:56:47 AM6/18/18

to Prasun Anand, John Woods, Sameer Deshmukh, Pjotr Prins, Pranav Garg, SciRuby Mailing List

Haha.....

I know that.

There's no way, I am getting a lot of traffic but Ruby still doesn't have any decent plotting library.

Maybe, some people might like it, I don't know.

Perhaps I am a little too romantic about keeping my blog pristine.

Regards,

Arafat

Pjotr Prins

unread,

Jun 18, 2018, 3:17:45 AM6/18/18

to Arafat Khan, Prasun Anand, John Woods, Sameer Deshmukh, Pranav Garg, SciRuby Mailing List

The agreement is two blogs per month.

> [4]https://groups.google.com/forum/#!topic/sciruby-dev/x3xlGKcCCK4

> To add to this, this helps other students and researchers get involved
> with the project.
> Regards,
> Prasun
>
> On Mon, Jun 18, 2018 at 6:24 AM Arafat Khan
> <[5]arafat....@gmail.com> wrote:
>
> On Sun, Jun 17, 2018 at 5:29 PM Arafat Khan
> <[6]arafat....@gmail.com> wrote:
>
> Hi Prasun,
> I have mentioned before in my emails, that I don't have any issues
> writing a blog post.
> I just don't want to write a blog post for the sake of it. I am happy
> to document my progress and discuss it among ourselves in elaborate
> emails.
> Usually, I use my blog to introduce software and write interesting

> ideas to build similar things( [7]https://medium.com/@Arafat./ ). I

> image.png
> The colors are kinda messed up right now but that can be taken care
> of..
>
> On Thu, Jun 14, 2018 at 12:13 PM John Woods
> <[13]john.o...@gmail.com> wrote:
>
> Want to send us a screenshot?
>
> On Thu, Jun 14, 2018 at 10:47 AM Arafat Khan
> <[14]arafat....@gmail.com> wrote:
>
> Hi Everyone,
> I am working on bar charts right now and things are looking better than
> yesterday.
> I am able to draw the rectangular bars and still working on adjusting
> the spacing in the bars to make things work well. I hope I am able to
> finish this by the end of the week. Currently, I haven't focussed on
> the data normalization aspect but that would certainly influence the
> outlook of the graph. So first, I am just looking at making the
> rectangle bars and focusing on spacing etc on some test data.
> Regards,
> Arafat

> On Wed, Jun 13, 2018 at 8:52 PM John Woods <[15]john.o...@gmail.com>

> [17]https://github.com/matplotlib/pytest-mpl.

> Right now this week, I am mainly focussing on making bar charts.
> Regards,
> Arafat
>

Arafat Khan

unread,

Jun 18, 2018, 3:22:59 AM6/18/18

to Pjotr Prins, Prasun Anand, John Woods, Sameer Deshmukh, Pranav Garg, SciRuby Mailing List

Okay Everyone,

I will write 2 blogs by this week and take care of the agreement with time. Besides, I really like writing them too as it will serve as a good reference.

Thanks

Regards,

Arafat

Sameer Deshmukh

unread,

Jun 18, 2018, 8:46:52 AM6/18/18

to Pjotr Prins, Arafat Khan, Prasun Anand, John Woods, Pranav Garg, SciRuby Mailing List

Also, you don't _have_ to write one of those fancy medium blog posts
with the cool picture on top. For the science community in general,
blogs are a way of communicating progress in a project. When someone
wants to view the decisions made on the way to implementing the library,
they will read the blog posts.

If you're too embarrassed to put it on medium, send it to the sciruby
blog on github. We love such stuff.

Reply all

Reply to author

Forward