Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Variance

variancecharts.com

Post reply on HN
Print view View original

41–50 of 56 posts

Re: Variance

#41

Post by throwawy »

The example graphs confuse dependent and independent variables, putting quarters on the vertical axis. This makes the curve look negative, when in fact it's positive (up and to the right - usually a good thing). This really takes away from the example, as it takes seemingly important business data, and renders it in a negative way that I wouldn't expect or like. I strongly suggest correcting the example.

Otherwise this looks awesome.

P.S. Should this be a Show HN? It feels like one.

Re: Variance

#42

Post by pwang »

jmduke wrote: post #9

This reminds me a lot of Vega ( https://github.com/trifacta/vega ), albeit using a HTML/CSS approach as opposed to a JSON one. By default, I'm excited for new ways to approach data visualization, since its a really exciting topic with no clear winners. I was worried about the hefty price tag until I realized that it was free for non-commercial projects. Will definitely try and play around with it this weekend, as I i…

Have you played with Bokeh at all? http://bokeh.pydata.org

Re: Variance

#43

Post by anigbrowl »

glaugh wrote: post #4

Apologies that this is about form not substance: I appreciate the clarity of the "What makes Variance unique?" and "Where is Variance not a good fit?" sections. Really useful. As someone who makes a data analysis tool (with a bit of viz thrown in), that's probably something we could do better.

I came here to praise 'Where is Variance not a good fit?' as well. This is a great introduction, I wish everything had this level of detail on the landing page. I find it infuriating (everywhere from Unix command line to websites for new tools ) not to have a clear answer to the question of 'why you should/should not use this.'

Re: Variance

#44

Post by jmduke »

pwang wrote: post #42
jmduke wrote: post #9

This reminds me a lot of Vega ( https://github.com/trifacta/vega ), albeit using a HTML/CSS approach as opposed to a JSON one. By default, I'm excited for new ways to approach data visualization, since its a really exciting topic with no clear winners. I was worried about the hefty price tag until I realized that it was free for non-commercial projects. Will definitely try and play around with it this weekend, as I i…

Have you played with Bokeh at all? http://bokeh.pydata.org

I have played with Bokeh! The capabilities seem really impressive but my main problem with it is its lack of embeddability and flexibility, which I imagine will get ironed out as it matures (though the Python bindings are on point).

Re: Variance

#45

Post by jwr »

glaugh wrote: post #4

Apologies that this is about form not substance: I appreciate the clarity of the "What makes Variance unique?" and "Where is Variance not a good fit?" sections. Really useful. As someone who makes a data analysis tool (with a bit of viz thrown in), that's probably something we could do better.

I also admire the upfront and clear explanation. For people who are about to make a decision this helps immensely.

Everyone should learn from this: there is no point in overhyping or hiding things — I will discover these tradeoffs eventually anyway, so why not save everyone some time and clarify them upfront?

Re: Variance

#46

Post by jwr »

lynaghk wrote: post #7
vrepsys wrote: post #5

truly nice looking charts. Wondering, if the colors and other styles built-in or the user has to define them? Also, seems only IE10+ version is supported, wondering what prevents the library from working on IE9, since (I think) svg is supported on v9 as well.

Variance is production-oriented which, for data visualization, requires a ton of the "designer interface" as much as it does from the "developer/analyst interface". The CSS is broken out into necessary structural bits (where everything is gray, there are no margins, &c.) and then into visual themes. If you aren't a designer, you can just use the stock theme. If you are a designer then you can start from the ground up…

I think what was meant was whether Variance comes with built-in themes. At least that was my first question — having a set of well-designed themes would add a lot of value.

Re: Variance

#47

Post by cheald »

diminoten wrote: post #21

Why are all the good chart generators that aren't dwarf fortress-esque always for-profit ventures?

Probably because someone put in a lot of effort to write visualization software that isn't dwarf fortress-esque and would like to be compensated for it.

If you'd like to build an equivalent open-source library, though, you have my vote.

Re: Variance

#48

Post by gburnett »

glaugh wrote: post #4

Apologies that this is about form not substance: I appreciate the clarity of the "What makes Variance unique?" and "Where is Variance not a good fit?" sections. Really useful. As someone who makes a data analysis tool (with a bit of viz thrown in), that's probably something we could do better.

The "Where is Variance not a good fit?" section is useful. I'm not just saying that as a long time victim/critic of pie charts.

Re: Variance

#49

Post by osi »

lynaghk wrote: post #3
osi wrote: post #2

nice. looks to be built on d3.

Hi, developer lead here. Variance is not built on D3; we're using some of the scale code, but that's it. In fact, variance-standalone.min.js is about 10kB smaller compared to current d3.min.js = )

ah i stand corrected! when looking at the minimized code some looked very similar to d3's minimized code. either way, impressive stuff!

Re: Variance

#50

Post by sdoering »

Having to load the "free" version from the variance cdn is the showstopper here.

I really refrain from sending my users data to anyone else, that is without my control. So sadly, as beautiful, as I find this, I have to refrain from using it in my non profit projects.

Post reply on HN