₦airaland Forum

wisemania:
One quick question,

how many of us think the requirement listed by various recruiters(Nigeria) for data Analyst role is quite annoying?

I mean,

....
Should be adept at queries,
Should carryout db maintenance regularly (why the hell do they have a DBA),
Should install and update systems software (DBA job),
Should carryout software tests(Really?),
Systems design and maintenance (Data Engineers/ Data Architect role),
Machine learning (data science ),
Experience with Google spreadsheet, GA, Qlikview, python, JavaScript, C#, tableau, PowerBI, etc (Alright)
Etc........


Men, I'm exhausted already!!!!

Make system no go blind my eyes!

Olalekank:
I found this online

Some companies abroad that can sponsor your visa - https://github.com/shubheksha/companies-sponsoring-visas
The list keeps growing

cochtrane:
I think many data scientists/enthusiasts will agree that SQL is the foundation of data analysis. In fact, some jobs require you to know SQL and nothing more. If you are getting to data science and don't have knowledge of SQL, then you've still got a long way to go.

For those struggling to learn SQL, I've got some news for you. I'll be organizing a SQL class for free in some days (maybe a few weeks?). If you are interested, send me a PM and I'll add your name to the list. When we get to a reasonable number, I'll open the class. Perhaps we can get new people interested in SQL, or help beginners firm up their knowledge.

Mode of delivery: online (zoom)
Time: will be about 7- 9pm Nigerian time on a weekend.
Resources needed: A laptop and good internet connection
What to expect: You will be doing some coding yourself, while following the lecture

We will use SQLite. But most of what will be learnt can be extended to MySQL and Postgres.
Shoot me a PM if interested. Hopefully we can get some more engagement going here for enthusiasts!

If we get enough interested candidates, we'll open the class.

Glokit:
Top schools to study... on average budget

Griffith College
Technically University Dublin (Previously DIT)
National College of Ireland
Dublin Business school

For your interest... its better to study in Dublin although the standard of living is very expensive but you will have more access to jobs compare to any other city...

For jobs check
indeed.ie
Jobs.ie
Jobbio.com
Glassdoor.com

For students jobs and side Hustle it’s better to enroll wit agencies like
Noel Recruitment
Excel Recruitment
Cpl Recruitment
Temple Recruitment
First staff Recruitment
Addecco

For accommodation
Datf.ie
Rent.ie

For food/upkeep
Lidl
Aldl
Tesco

More Updates soon... but u can ask questions

cochtrane:
This is an interesting question, and this is how I'd do it.

Prior probability that the woman has cancer is P(A) = 0.01
Let event A be that the woman has cancer
Let event B be that the woman gets a positive mammogram test
So, we are looking for the probability that the woman has cancer given that she gets a positive mammogram test, i.e P(A|B)

Using Bayes theorem, P(A|B) = (P(B|A) * P(A))/(P(B))
We know the prior, which is P(A).
P(B) can be expressed as P(B|A) * P(A) + P(B|not A) * P(not A)
P(not A) = 1 - P(A) = 0.99
We take P(A|B) to be 0.75, which is what the physicians claimed, i.e. the probability that the woman has cancer given that she has a positive mammogram test.

P(B|A) is the probability that she gets a positive mammogram test, given that she has cancer. That probability is 0.8, since the mammogram accurately classifies 80% of cases.

If you substitute everything into the equation, you get P(B|not A) = 0.27%. This means the probability that the woman gets a positive mammogram result given that she does not have cancer is 0.27%. This is the size of the false positives and honestly is too low. This would suggest that the test is incredibly accurate, which as we can see from the figures of 80% and 90%, is not the case. So, it is very unlikely that the 75% from the physicians is correct.
I suspect the "benign" case is the complement of the "cancerous" case, though it's not clear. If that's the case, then the probability of getting a negative test given that she does not have cancer (it's benign) is P(not B|not A) = 0.90, so that P(B| not A) = 1-0.90=0.1 or 10%. This is more likely the value of your false positives. If we compare with our earlier 2.7%, we see it is much higher. With this we can confirm that the physicians obviously are not right in their diagnosis.

Olalekank:
https://twitter.com/wejapahq/status/1296341805752082433?s=21

Check out WeJapa on Twitter
Hiring for Amazon Web Services - 8 roles!

wisemania:
We need to be careful when dishing advice.

We get it! You are a bookworm, you can't help it, we understand, but be careful with your advice. Ejiod already explained the process of becoming a data scientist.

How?
By FIRST becoming a data analyst. You do not need to understand the intricacies of advanced statistics (or ML) to become a data analyst. What you need is

1. basic mathematics (statistics),

2. basic intuition of how ML algos work(they've been written in packages already for your usage). Kiril eremenko did a great job on that on udemy. Just get the course for free and learn the intuition with how and when to use it.

3. Excel,PowerBI(because it's cheaper for companies to integrate as most companies rely on Microsoft products and it's simply advanced excel) and or Tableau (or any other visualization tools). Tableau is a bit expensive though

4. SQL (T-Sql/MySQL/P-SQL) they are all 90% the same. Learn MySQL and the rest will become a breeze!

5. A programming language (Python/R/Julia, etc). You don't need to know all the functions of the language, just learn the basics and the necessary libraries for now (trust me, there'll be plenty of time to dive deeper into these languages later!)

After you're pretty solid on the above, start applying for jobs (junior data analyst or Business Intelligence Analyst).



If you need to feed, get a job with excel,powerBI/tableau,SQL,Python/R
then learn the remain (Statistics, ML-DL) on the job as required.

Else, you go HUNGRY


This is barely my [b]3rd month into data analysis and I've gotten some interviews already... No time to waste, as you need to survive...
[/b]

Olalekank:
Been a little cold here

Some companies that might be hiring
https://careers.bloomberg.com/job/search
https://www.palantir.com/careers/
https://stripe.com/jobs

Here’s a link to some companies that might be hiring abroad. It’s for new grads but you can check out their careers page
https://github.com/cmackenzie1/new-grads-2020


Here’s a Leetcode question to warm guys up a little
https://leetcode.com/problems/add-two-numbers/

emperordelis:
Mine was sometime around 2017,i had just completed my professional exams and it was only natural i start looking for a better job that befits my new qualification..Rearranged my CV and started bombarding marine / shipping companies and their agents with my CV. After about a week, i got a mail from one of them asking me to be in Port Harcourt the next day for an interview. I was excited,couldn't even sleep well,left Uyo very early the next day and got to PH in time..Went straight to the address she asked me to come to,if i had known how this day would turn out to, i probably wouldn't have wasted my time going there.
I got to the gate,security man won't grant me access without getting approval from whoever asked me to come. I spent over 30 minutes trying to call this lady,no response even after the security man assured me that she was inside,decided to call one last time and call it quits,she eventually picked up and gave me the worst attitude / insult i have gotten from any HR Personnel so far. Insulted me for calling her number,i tried to make her understand that i didn't come on my own,just yesterday she sent me an email asking be to come for an interview. I have never felt so embarrassed in my entire life,i apologised for calling,ended the call and cursed her to suffer for 6000 years..Wasted transport,wasted energy,wasted false hope just to come and receive insult... Some HR people totally forget the Human in their Job title

Ayinke93:
I use Udeler to download my Udemy courses on my Laptop. And you can get torrent files from Freeturialsusa.com then use uTorrent or BitTorrent to download them.

HewlettPackard:
Nice. Anyway of downloading the videos for offline use?

Modified: got it, service avialiable on the app.

ibromodzi:
Modified:
Steps in determining what tests to use in a statistical analysis;
1. What type of variables are you dealing with? Numerical or categorical?

2. What type of analysis are you doing?
a. comparison (mean,median,proportion); this is where you use t-tests, ANOVA, etc.. depending on the distribution of the data(parametric or not)
b. Relationship between two variables (say gender and smartness); this is where you use correlation (Pearson or Spearman rank)
c. Predicting one variable from another (say exposure to smoke predicts the risk of lung cancer); this is a Regression task

3. Number of groups involved (say the effect of Chloroquine in Covid 19 and non Covid 19 patients - 2 groups)

4. Distribution of your data (Normal or not); there are different ways to determine this.

I believe you now understand the point I was talking from when I mentioned theoretical knowledge of statistics.

KlausMichaelson:
Wow, I haven't gotten to that level then. Thanks for the insight.
Do you mean you can do a computation on a new column that requires values from other columns at just a single input? And it will also give answers to every cell in the new column? At just one single input or formula??

KlausMichaelson:
Good day Sir to answer your question below:

**But i observed that what is being done using Pandas and Matplolib can be done comfortably with excel(Power Queries)**

As an intermediate, I'll like to let you know why pandas is much more flexible than Excel.
Here are my reasons

(1). Using Pandas, you can easily create a new cell(or column) that could do a multiple computations(addition, Subtraction, formulae etc) using values from other cells(or columns) in just a single line of code.
But Excel is limited to that as you can only
do it for each cell in a new column one by one and this can very cumbersome.

(2). Excel doesn't help give you a visualization of your preferred taste. That doesn't mean you can't design a plot( it's axis, Gradient line, ticks and the likes). But Pandas is more flexible using Matplotlib and pyplot, you can easily give your plot a good visualization of your choice.You can easily change marker colour, line size, line colour and the likes by just a simple line of code(even shorthand notation)

There are other reasons why I think Pandas/Matplotlib/Pyplot is more flexible than Excel but those are the ones I can remember at the moment. Although I stand to be corrected cos I'm still a learner.

Our Pros in the house can help too

tplayer:
1- Excel(and relatives/natives like pivot table, power query(you can skip the M language in it), learn to create dashboards with excel(because what a lot of business people know is excel and excel is everywhere; power bi is d shiny kid getting lots of love now, and for good reasons), learn formulas and array formulas. Once u understand how cell referencing works, excel becomes easy.
1.5- power pivot (or just move to 2)
2-power bi(please learn data modelling here, please please learn how to use d dax language to create calculated columns, measures, time intelligence so you can stand out. power query and power pivot are still here too.
Note that tableau can replace power bi too. These 2 are d most popular.
3. - involve urself in d statistical aspect of what you are learning. Using tools without knowing d statistics is not complete. after all data science, forger d sexy name, is basically statistics, some pure mathematics, cimputer science, and common sense. Please read all you can about probability, distributions, Central limit theorem, experimental design, ANOVA, manova, correlation, regression, Exploratory data analysis(e.g make sure you know all about d box plot, not just how to plot it), parametric and non parametric test, post-hoc tests. etc. Baba if you try all d things here, u are already a small boss.
4- Python or R, you choice (to implement 3, and even parts of 1 and 2. They are d languages of data.

5-Now, advance statistics and calculus( For MACHINE LEARNING!!!!!! which can be implemented using python or R.) You can start machine learning by using clicks and libraries before you go under d hood. machine learning no be beans sha o and its not just simple regression. But its learnable.

6. this has no number really. insert SQL anywhere after 2.


7- PRACTICE. PRACTICE. PRACTICE!!!

If all you want is mainly extraction, cleaning, munging, visualisation of data, and creating a shiny(apologies to R guys) dashboard, A simple knowledge of statistics may do it. Excel, Power bi and SQL will do.

To become a real data scientist aka machine expert, statistics, maths, must come in.

please, I put dis together in a hurry. others here should please chime in.

electromagnito:
interesting question, well firstly, I had to pray to God for guidance because its all by his grace I find myself in this position.
I had wanted to go to Hong Kong no matter what because the PhD topic was more related to my masters thesis which was on Glass science plus Hong Kong Universities are all in Top100 in the world including were I was admitted, let just say if there was no covid-19, I would have been in Hong Kong by now definitely.
During the lock down, I use the opportunity to also apply for a PhD at University of Hawai'i in US to work on bulk material of colloid particles, this still fall under my research work but more of biophysics, I got admitted after two Skype interview and I believe my recommendation letter from my two professor in France played a bigger role, I was told, I was selected based on my molecular dynamics simulation skills(trust me I learnt everything in france, Nigeria na scam), but there is one problem, covid 19, so I can only resume in US for spring next year plus I will be fully funded by American Chemical Society, but after going through the journal given to me to prepare myself for spring, baba I got more confused.
For France, La France, i got a PhD with ENS Lyon which is equivalent to getting a PhD with MIT in US, the PhD will be undertaken at Ecole Mines though but one special interest here is the topic, very interesting and fully sponsored by EDF, I will be working on nuclear reactor microstructure material like the 316L steel screw, this is a first hand industrial experience with EDF, everyone wants to work with EDF for sure.
I believe i will go to US someday for postdoc probably, but I won't kill myself for a US PhD because I am desperate. When the time comes, we will move trust the process baba