New role desired, data analyst/data visualisation type - how to get there

413x

413x

413x

Caporegime
Joined
13 Jan 2010
Posts
33,138
Location
Llaneirwg
My current job isn't going anywhere. The time line for potential challenges of interest is just too long and to be honest I don't have faith it will ever get there.
There is also potential that this job will end 12 months from now due to external contract renewal.
I have virtually no social circle, do not like the area, and am only here for said job

In looking for advice on what to do next as in the last couple of years I've gained a lot of skills. But only a few are concrete.

Basically I have a broad understanding of statistics and some programing

Very competent in vba
Partially in sql
Partially in R
And started learning python
Also familiar with Qlikview BI tool.

What I really like is visualisation and creating tools. I'm happy doing most complex code based jobs but really mainly around data as opposed to pure coding.

From looking around a safe job (considering the 12 month time line) would be a vba/sql/BI type role with R/python as desirable

Main issue here is sql as I'm only able to do basics
I'm talking joins, creating tables, creating schema, queries, loading data in


What is the best/quickest way to improve sql? Without any real need for it in my current role, what can I work on (ie project) to help here?
Is the best idea to improve sql over other options?
I have mysql in my laptop and try to do any data work in it (loading in csv, using odbc to link to R, Excel, qlikview, other)

There seem to be a fair few 35k+ London roles which state vba/sql + other.

This might well be effectively a pay cut as I'm on 31 in Peterborough, but I am of the opinion either in going to go crazy where I work due to lack of challenges or find myself redundant come 12 months
 
Last edited:
How good are your qlikview skills?

I don't rate your chances getting a sql job without real work experience to back up your knowledge. When I went for sql jobs in the past I always had to sit a test or verbally tested by a techie.
 
You don't need advanced level SQL to work in Analytics, depending on the type of job. Good query skills coupled with a bit of ingenuity and Google (for the stuff you don't know) can be fine when coupled with other skills you have mentioned such as R and visualistion tools like Qlikview. And being a data analyst isn't just about technical/language skills, it is also about being able to apply a logical mindset to work through a problem or unexpected scenario in the data.

If your background is in Stats then I would definitely look at firming up your knowledge of R, from what I can tell there is a trend moving towards it and with Microsoft pushing it fairly heavily in SQL2016 (Revolution R) I can see it being something that could see you stand out from the crowd. Python another option too.

That said, recent surveys have suggested SQL is the most in demand language of all.

£35k is quite pessimistic for a data science job in London. I would say £45k+ but might depend on how transferable your current experience is.
 
I think there is a lot of variation in the requirements for roles... 'data analyst' and 'data scientist' are ill defined and can mean anything from spreadsheet monkey through to roles requiring post grad level statistical knowledge and a whole range of stuff in between.
 
i guess im half way between.
im used to statistics (regressions, univariate, multivariate, random forests etc)
im also known as a bit of a realist in work. as quite often have to tell people what they think they have isnt statistically correct. i know a fair bit around real life stats (ie, 0.05 is an arbitrary figure that is only 'significant' as it it is convention).

i do not consider myself a data scientist. i cant program a neural net, but i have written a basic image analysis program in python. however i do not know R/Python well enough.
this is mainly as i learnt my stats in a GUI not code based application. (a SAS product)

and this is the role im looking for. one where i can develop these.

sql, i can query fine, and if i cant, it doesnt take me long to do it.

i consider myself half way between data scientist and data analyst.
i am certainly not a graph builder


im pretty good with logic, graphics and dash-boarding (some of the simple ones i have made within the first month of me using qlikview went down well).
presenting stats to non stats people is probably one of my valued skills. be it grapics or in person presentations


if i can really get into it, i know the salary expectations can get fairly substantial
 
Last edited:
fretted said:
A data analyst is not the same as a data scientist!
Click to expand...

can be... depends on the role itself - they're ill defined titles, it isn't like there is some regulator or professional body that had laid out the requirements for someone to call themselves one or the other - different firms have very different requirements
 
suggestions re: the above:

the University of Sheffield offers an online MSc in Statistics:

http://www.sheffield.ac.uk/postgraduate/taught/courses/purescience/mathstats/statistics-msc

Columbia offers an online MS in Data Science

http://datascience.columbia.edu/master-of-science-in-data-science

I'd suspect that going forwards you'll see more online/part time courses being developed for this area.

Also, if you do move to London, then Birkbeck offers a part time MSc in Applied Statistics:

http://www.bbk.ac.uk/ems/courses/msc_pgdip/msc_statistics
 
Last edited:
Ooft, don't think I could afford those courses.
The UK one would take 3 years at nearly 3k per year. (assuming cost over 3 years is 8000/3)
If I had that much of could commit to education I probably wouldn't need the qualification itself.
 
you can take out a govt loan, they've just been made available this year for post grad taught courses:

https://www.gov.uk/funding-for-postgraduate-study

:)

That Birkbeck MSc course is just under £4k for the whole course... so 2k per year... would be quite useful if you moved to London for a junior data related role and wanted to progress to one requiring better stats background - such as the 'data science' type jobs... Govt loan covers up to 10k so more than enough. could well be worth the risk if you land say a junior role and get other skills up to scratch given the jump in pay for the more quantitative roles. It isn't an easy option, but then again this sort of stuff almost never is.
 
Last edited:
That's much more reasonable. And might not even need a loan for it (although the loan maybe useful)
And that sounds like a reasonable route + shows drive to improve etc.

If I can get a junior role with OK pay with the opportunities that come with London might not be a bad ideal.

Thanks dowie
 
no worries, I'm interested in the same area myself and looked at various options a while ago

one last thing - you didn't mention undergrad degree I'm assuming you do have a quantitative undergrad (as you've mentioned stats knowledge) but if you don't then it looks like Birckbeck also have a part time 1 year course to bring you up to speed with the undergrad maths/stats needed for a Stats Masters (as it seems it is a proper stats MSc not a conversion course)

http://www.bbk.ac.uk/study/2016/postgraduate/programmes/GDCSTATI_C
 
I read that
I have a molecular science degree which sometimes it's enough with work experience but sometimes isn't.
Holes in my maths are things like matrix algebra.
But a whole extra year? That sounds rough. As much for the time as the money.
I would seriously need to consider if 3 years study is worth it vs carrying on with work and free time learning (coursera foe example)

Also would I have the stamina for 3 years?
I was always good but never gifted at maths. Particularly pure and any you can't conceptualise.
Really, my true calling was engineering. Physical/mechanical maths, real world tangible concepts.
However at 30 there's no way really to get there.
 
Last edited:
Well how much of a time drain it is presumably depends on how much you don't know vs how much is revision from stuff you've self studied

It isn't pure maths, it looks like one applied maths course (covering linear algebra, ODEs, PDEs - basically the same maths that would be core to plenty of engineering degrees too) and stats(hypothesis testing, ANOVA, regression etc..). if you're not familiar with that stuff then a stats MSc is a bit of a non-starter anyway as you need good linear algebra skills to study stats properly at an undergrad level let alone post grad.

For a 'proper' data science role(granted it is still an ill defined term that often overlaps) ti seems this sort of stuff is often required on the other hand it seem the roles more geared towards 'data analyst' don't necessarily need it and simply knowing how to apply certain things and how to use various tools, languages is sufficient.

Maybe it is worth seeing what sort of role you can land right now and whether you need to up grade your skill set/knowledge if trying for some of the higher paying roles later down the line. Maybe some udcaity/coursera type courses might be a good/quick way to fill any gaps you're concerned about in the short term. :)
 
I think personally I'd rather have to do more work (ie catching up) if possible rather than make it 2 years
It's hard to know if I know enough maths from work and reading around problems that doing the foundation course is worth it. Probably worth contacting them if indeed I do get to London.
 
you could get a good idea of that from the syllabus details of the two modules listed on that course

if you get yourself a copy of advanced engineering mathematics by Stroud or alternatively by Kreyszig then you'll find most of the undergrad mathematics you'll cover in an Engineering degree and/or the applied part of the first + second year of a mathematics degree. Looking at the syllabus pretty much most things from that mathematical methods module is covered by either of those books (well aside from markov chains).

Both books contain chapters on probability and statistics but those are only partly sufficient, you'll likely need an undergraduate stats and probability book too.

There aren't really short cuts here, you either know this stuff or you don't - if you know most of it then you could see about getting straight onto the MSc there or elsewhere. If you don't know it then it might well take several months to a year of self study anyway to get to the point where you're ready to start an MSc.

Alternatively there might be some MSc courses with looser entry requirements - but then the downside is you'll end up with an MSc that mostly covered undergraduate material which is probably not sufficient for some roles.
 
Last edited:
Yep I know. If I can't do it I can't.
There are probably too many holes to jump into advanced stats, but paying for a years foundation where I may well know 3/4 of it also seems like a waste.

I'm going to put my cv out with
SQL
VBA
BI tool (qlikview)
And also that I'm new to R/python but won't apply for jobs requiring them.

To be honest I'm not wanting the world salary wise. I'm more looking for something I won't struggle on and more importantly, I can learn on.

I've seen some that appear to fit the requirements.
What I want is a role that is a bit stats and a bit BI.
 
You must log in or register to reply here.
Back
Top Bottom