London Clojurians monthly gathering tonight at SkillsMatter from 6.30pm onwards. https://skillsmatter.com/meetups/10755-london-clojurians-monthly-gathering

This month we have Steve Smith giving a talk about how Atlassian use Clojure (and Clojure job opportunities there). I'll also talk about ClojureX, our annual conference we have just started to Organise.

Morning

I can also give a recap of the ClojureBridge London event last weekend.

Sounds like a very interesting talk @jr0cket

The London Java Community is getting interested in Clojure again. We had a great virtual Meet A Mentor session where we discussed the value of learning and developing with Clojure

moin moin

morning

@jr0cket if you can move the meeting to 8 May I could give a talk on "Dealing with Annoying Sized Data using transducers and core.async"

I'll not be in London on 1 May tho

morning

Morning.

morning 😄

Morning All 🙂

@otfrom - Sounds like a talk I would benefit from...

@maleghast me too. That's why I want to do it. 😉

any suggestions on what dataset I should work with? I was going to go back to my old standby of NHS GP prescription data

Hehe

something in the 10s of GB range would be good

Well, I find the CRU archive to be of an annoying size

CRU?

Climate Research Unit (UEA)

link?

hold on I'll get you a link

thx 🙂

http://www.cru.uea.ac.uk/data It may not be big enough for your needs however

I am in the process of working out how to "harvest" large tracts of NOAA data (GHCND datasets for Non-US countries) so I am sure that I can point you at much bigger datasets if you want to play with climate info, but there are "snags" in terms of how big they are and the limits on download / access...

It’s a shame that the UK pub names dataset isn’t big enough for you 😛 - https://www.kaggle.com/getthedata/open-pubs

yeah, there is around 88GB of NHS data in csv format

Crikey

That's a LOT of CSV

for the prescribing data anyway

and I know how the dataset is put together

Probably best to use that then...

Plus the CRU archive is not CSV, it's fixed width columnar text, so it's "fiddly"

and you have to scrape it page by page...

Honestly I don't know what I was thinking...

Sorry 😞

I'd love to find come dine with me data.

apparently there's a strictly come dancing data set.

There's an open dataset for Strictly?

*boggle*

https://dataissexy.wordpress.com/2016/10/18/calculating-the-darcey-coefficient-part-3-strictlycomedancing-machinelearning-clojure-weka/ from our very own @jasonbell

That should be https://dataissexy.wordpress.com/2016/10/16/calculating-the-darcey-coefficient-part-1-strictlycomedancing-bbcstrictly-darceyofficial-numbers-clojure/

http://www.ultimatestrictly.com/downloads/

I ❤️ @jasonbell

Thanks @dominicm - very elightening 🙂

Well there's one citation I wasn't expecting. Thank you @maleghast @dominicm 🙂

Hehe

@jasonbell - I am going to "follow along" with your posts and do the weka one at least, when I'm not working - looks fascinating on first read. 🙂

I can't remember if I pushed that stuff up to github or not. I'll check this evening

@maleghast https://github.com/jasebell/dareceybussellneuralnetwork

It's been a long while since I ran it so let me know if you bump in to any problems.

Thanks @jasonbell - will do 🙂

👍

Does anyone know if there is a way to take a Terraform file and process / parse it into AWS cloudformation?

@maleghast first kill a:chicken: at 🌕

@mccraigmccraig - Lol!

@maleghast I think @mccraigmccraig is reading from the official documentation... 😉

@maleghast cloudformation is the work of 😈

@mccraigmccraig - Yes indeed, I don't really disagree, however rather than have a non-standard install of Terraform on one of my instances I would like to use Terraform to output "correct" Cloud Formation so that I can use it in conjunction with the AWS CLI to create infra.

i.e. I have Terraform on my workstation, I perfect the "recipe" in Terraform, test it and then when I want to set up a remote machine to periodically create / destroy infra (on a schedule) then I export the CF from Terraform (if that's possible) and put the CF onto the instance that is going to be driving...

And yes, this would all be a lot simpler if I had a cluster to run workloads in / with Docker, but the intention at the moment is to limit costs and an always-on cluster is a lot more costly than an instance that gets spun up once a week for a couple of hours by one of the very small number of instances that we are already running 24/7

Google leaves me thinking that Terraform -> CF is not possible, but I coulda sworn that it used to be...

(I also am pretty certain that the AWS provider for Terraform is in fact taking the platform agnostic Terraform file(s) and creating CF in order to work)

i don't think the terraform AWS provider creates CF @maleghast - i'm using terraform to provision our clusters and there are no stacks showing in our CF console after a cluster has been provisioned

Hmmmm... Maybe it's just turned immediately into AWS CLI commands..?

I am not happy about it, but installing terraform directly from the HashiCorp website onto the instance that will be using it is pretty pain free.

Now all I have to do is bend my Jenkins server to my will...

Can't get it to interact with Github at the moment, it's being all "Jenkins" at me.#

Terraform doesn't create cloudformation. It does it's own normalization. It uses the AWS APIs (maybe the go SDK?) however, terraform can manage cloudformation stacks, if you need to mix and match.