If you ask any data scientist, they will probably tell you 90% of their time is spent on data processing/munging.
The success of your analytics results, insights, and quality of your model depends on the quality of your data.
Take a machine learning modeling project, for example:
The overall data process from raw data to clean, ready-to-use data usually involves the following steps:
- Data acquisition.
- Talking to domain experts and identify the source of the data, understand how the data is generated, if it is of high quality (machine-generated vs. manually entered);
- Data Preprocessing
- Remove or impute missing data, extract features from textual or categorical data, normalize some data, split the data into training vs. testing, down/upsampling, etc.
- Data Postprocessing
- Sanity check to make sure there are no apparent mistakes were introduced in previous steps;
- Remove outliers or special cases;
And you will likely need to use SQL in every single step!
Now you are convinced SQL is essential for your data science career, how about start learning on sqlpad today?
