Ask HN: How to practice data analytics skills?
8 comments
From a product analytics perspective I’d suggest :
1. Think of a product that exists. Define a goal for the product and success metrics you will use. Dau, mau, user retention,incremental revenue etc. 2. Come up with gaurd rail metrics 3. Define performance and reliability metrics as well.
Now try to figure out how you would construct queries to answer these questions. And how would you visualize this info.
Then if you can find datasets or create synthetic data sets to actually write these queries or better yet create pipelines that ultimately feed a dashboard I think would be worthwhile.
1. Think of a product that exists. Define a goal for the product and success metrics you will use. Dau, mau, user retention,incremental revenue etc. 2. Come up with gaurd rail metrics 3. Define performance and reliability metrics as well.
Now try to figure out how you would construct queries to answer these questions. And how would you visualize this info.
Then if you can find datasets or create synthetic data sets to actually write these queries or better yet create pipelines that ultimately feed a dashboard I think would be worthwhile.
Merry Christmas - Have you tried https://www.kaggle.com/datasets ?
I would start with a familiar dataset and then see if you can prove your intuitions with the data. For example, if you use a smartwatch you can download your sleep data and check if it supports the hypothesis that during weekends you go to sleep later. Then, you can also look for other insights and then check if they are compatible with your prior hypothesis.
On the off chance that you're into sports, my company StatsBomb have some free data for both soccer and American Football up on GitHub:
https://github.com/statsbomb/open-data
https://github.com/statsbomb/amf-open-data
https://github.com/statsbomb/open-data
https://github.com/statsbomb/amf-open-data
Data Is Plural is a weekly newsletter (and seasonal podcast) of useful/curious datasets, published by Jeremy Singer-Vine. There have been 356 editions, dating from October 21, 2015 to December 20, 2023.
https://www.data-is-plural.com/
https://www.data-is-plural.com/
https://superset.datatest.ch/superset/dashboard/7/
Superset has a great dataset with prebuilt visualization for historical video game sales if that's interesting to you
Superset has a great dataset with prebuilt visualization for historical video game sales if that's interesting to you
Shameless plug, we're building exactly this: www.datawars.io
For now, we're hyperfocused on Python/Pandas/Scikit-learn as we're just getting stated (we launched in June). But we'll expand more tracks for data analytics and data engineering.
For now, we're hyperfocused on Python/Pandas/Scikit-learn as we're just getting stated (we launched in June). But we'll expand more tracks for data analytics and data engineering.
Merry Christmas buddy.
You'll find a ton of public datasets on GitHub [1].
Maven Analytics offers a monthly data analytics challenge [2] that you can enter for free. See their past competitions for some interesting datasets.
As I'm based in Ireland I'll also recommend the Irish Data Portal [3].
[1] https://github.com/awesomedata/awesome-public-datasets [2] https://mavenanalytics.io/challenges [3] https://data.gov.ie/
You'll find a ton of public datasets on GitHub [1].
Maven Analytics offers a monthly data analytics challenge [2] that you can enter for free. See their past competitions for some interesting datasets.
As I'm based in Ireland I'll also recommend the Irish Data Portal [3].
[1] https://github.com/awesomedata/awesome-public-datasets [2] https://mavenanalytics.io/challenges [3] https://data.gov.ie/
Not looking for books, but rather sample datasets I can use to visualize, analyze and test if I found the right insights.