These are chat archives for FreeCodeCamp/DataScience

8th
Nov 2017
Josh Goldberg
@GoldbergData
Nov 08 2017 00:00
@mcbarlowe thanks!
CamperBot
@camperbot
Nov 08 2017 00:00
goldbergdata sends brownie points to @mcbarlowe :sparkles: :thumbsup: :sparkles:
:cookie: 134 | @mcbarlowe |http://www.freecodecamp.com/mcbarlowe
Ghost
@ghost~580ed9c0d73408ce4f309ef0
Nov 08 2017 02:29
H
Matthew Barlowe
@mcbarlowe
Nov 08 2017 02:30
Hello
Ghost
@ghost~580ed9c0d73408ce4f309ef0
Nov 08 2017 02:31
Can I ask something
Josh Goldberg
@GoldbergData
Nov 08 2017 02:32
of course
Ghost
@ghost~580ed9c0d73408ce4f309ef0
Nov 08 2017 02:32
Can anyone contributed to world of open source
Does anyone know python here
Matthew Barlowe
@mcbarlowe
Nov 08 2017 02:35
I do
Ghost
@ghost~580ed9c0d73408ce4f309ef0
Nov 08 2017 02:36
In what and for which companies?
How to get started and get involved?
I am looking for help in my upcoming project. Does anyone can help ?
Matthew Barlowe
@mcbarlowe
Nov 08 2017 03:13
@vipuldcoder I do python mainly for data science I don’t work for any company I’m teaching myself right now if you’re looking to learn python I recommend www.automatetheboringstuff.com
Chikezie122
@Chikezie122
Nov 08 2017 03:37
@here good morning. Has anybody done anything in Natural Language Processing?
I have a thesis I am working on and I need help
Josh Goldberg
@GoldbergData
Nov 08 2017 04:13
@mcbarlowe nice book recommendation. I added it to my list. I’m currently learning R.
Chikezie122
@Chikezie122
Nov 08 2017 04:18
@evaristoc good day. I need someone who has done something in Natural Language Processing, specifically Terminology extraction. Do you know anyone?
Josh Goldberg
@GoldbergData
Nov 08 2017 05:18
just started playing with the data in R about an hour ago. @QuincyLarson. Seeing what I can come up with.
Timothy Javins
@timjavins
Nov 08 2017 05:29
@Chikezie122 We have a Gitter room for Python, too. You can try there. There is also an official Python tutor email list you can use. You can find it at python.org. Lastly, the Python community on Twitter is pretty helpful.
Matthew Barlowe
@mcbarlowe
Nov 08 2017 05:53
@GoldbergData I don’t know how advanced you are at R but the swirl package is a great built in learning material that you can run in rstudio
Josh Goldberg
@GoldbergData
Nov 08 2017 06:01
@mcbarlowe thanks. I’ve played around with swirl a little bit. It’s been a while. I’ll never be ‘advanced.’ Comes with the territory of programming. lol
CamperBot
@camperbot
Nov 08 2017 06:01
goldbergdata sends brownie points to @mcbarlowe :sparkles: :thumbsup: :sparkles:
:cookie: 135 | @mcbarlowe |http://www.freecodecamp.com/mcbarlowe
Matthew Barlowe
@mcbarlowe
Nov 08 2017 06:02
For sure I just used it a lot when I was learning r and they have some advanced statistical packages I haven’t done yet that look cool
sherlockdk
@sherlockdk
Nov 08 2017 06:10
@vipuldcoder i do
Josh Goldberg
@GoldbergData
Nov 08 2017 06:42
just barely got started. Only two visualization’s in, but sharing with you guys so far. @QuincyLarson
Here’s the git. I will have more time to work on it tomorrow. https://github.com/GoldbergData/Kaggle-Survey-2017. @QuincyLarson
evaristoc
@evaristoc
Nov 08 2017 09:46

Hi @Chikezie122

I think I am currently the most visible person in this group who has been working on NLP topics.

To be honest, it is the first time I hear the term "terminology extraction". However it seems like I have done some already :) .

I have been applying basic techniques for tag identification.

It seems also useful for creating vocabularies / glossaries, translations... useful in marketing. Also likely automated job search engines. Just to mention some.

Below some references:

Tell me your ideas and I will tell you mine ;) - I would prefer if you have done already some research, it shows you are seriously into the topic.

Ghost
@ghost~580ed9c0d73408ce4f309ef0
Nov 08 2017 15:49
hello
Alice Jiang
@becausealice2
Nov 08 2017 21:37
So I'm presenting a D3.js basics lecture on Saturday and haven't even started prep... any suggestions of what I should include?
evaristoc
@evaristoc
Nov 08 2017 21:48
@vipuldcoder show your projects and we will see if we can help?
@becausealice2 hmmm... no idea really... but if it is basic I would emphasise on the most important things to learn that would help to master D3.js? The cycle?
Alice Jiang
@becausealice2
Nov 08 2017 21:50
Cycle?
Matthew Barlowe
@mcbarlowe
Nov 08 2017 21:51
Lol d3 makes no sense to me and I’m actively trying to learn t right now
Alice Jiang
@becausealice2
Nov 08 2017 21:55
Which makes you my target demographic. What about D3 is confusing to you?
besides the fact that it's in JavaScript which nobody really understands?
Matthew Barlowe
@mcbarlowe
Nov 08 2017 22:03
The biggest thing is when anon functions use variables that haven’t been introduced before in the code I think the main one is d which means data I’m pretty sure
https://bl.ocks.org/mbostock/4060606 like this example here
The first argument fed to the method is the attribute name, and if you're setting and not getting, the second would be the value, right?
When the second argument is a function, D3 iterates over all selected elements (think data binding) and d is representative of each element which is being fed individually into the function
So in the choropleth of US counties, D would be each county's shape and expose all data bound to that shape to the function in question
does that make sense?
evaristoc
@evaristoc
Nov 08 2017 22:19
@becausealice2 the cycle or whatever : enter-update-exit or something like that.
Alice Jiang
@becausealice2
Nov 08 2017 22:19
Oh... Data binding?
@evaristoc this?
evaristoc
@evaristoc
Nov 08 2017 22:20
binding, updating and unbinding (exit)
It used to be the key aspect of d3.js before version 4. I don't know if it still is, time not using d3.js and haven't tried version 4 yet.
Alice Jiang
@becausealice2
Nov 08 2017 22:22
Data binding is still very much the key
evaristoc
@evaristoc
Nov 08 2017 22:22
What are the main things I will find different in version 4 @becausealice2 ?
Alice Jiang
@becausealice2
Nov 08 2017 22:22
method names, actually
They used to have things like `d3.scale.linear() but now it's d3.scaleLinear()
evaristoc
@evaristoc
Nov 08 2017 22:23
Yes... I remember some... they rebuilt the library to make a lot of methods more available I think...
They were talking about more modularity.
evaristoc
@evaristoc
Nov 08 2017 22:24
Yes but I don't want to read all that, I prefer to ask the expert...
Alice Jiang
@becausealice2
Nov 08 2017 22:25
So go find an expert
evaristoc
@evaristoc
Nov 08 2017 22:25
So it is why I am asking you...
:):):)
Alice Jiang
@becausealice2
Nov 08 2017 22:25
Ever since people started paying me to do D3, I've lost knowledge on it
evaristoc
@evaristoc
Nov 08 2017 22:25
I am not paying you so you don't have excuse...
Alice Jiang
@becausealice2
Nov 08 2017 22:26
And as soon as this internship is over I'm walking away from Visualization in general and moving to something less people will have to look at
I do have an excuse because my knowledge isn't conditional on who I'm talking to :joy:
I'm just super lucky there's at least one enthusiastic D3 developer in the D3 slack who has saved my ass so many times
evaristoc
@evaristoc
Nov 08 2017 22:30

And as soon as this internship is over I'm walking away from Visualization in general and moving to something less people will have to look at

Trying to find some ideas about what that could be but I give up. Not very creative tonight.

Alice Jiang
@becausealice2
Nov 08 2017 22:31
I'm going back to bare basics analysis, where the quality of the visuals matters only to me :)
evaristoc
@evaristoc
Nov 08 2017 22:32

All the things that I can think are very much visible, frankly speaking. Horrible...

Hey! I found something! Music!!!

Ahhhh.... :( I got it wrong...
Alice Jiang
@becausealice2
Nov 08 2017 22:33
I mean, analysis is visible, but the work emphasis is what the data says, not what it looks like
evaristoc
@evaristoc
Nov 08 2017 22:34
Yeah... if you are exploring data surely it is so...
evaristoc
@evaristoc
Nov 08 2017 22:42
It is a bit troubling when you have to use visualizations to communicate ideas to a public though. Then you have to pay attention to all the other details that you could easily ignore when you are analysing the data for your own interest.
Not necessarily hard, but a bit labourious IMO. Sometimes you need to know a bit more than just a few things in D3.js to do that well.

Anyway going to bed.

@GoldbergData Hope everything ok with the project! Again, keep asking here!

Alice Jiang
@becausealice2
Nov 08 2017 22:45
Sleep tight!