3 ms·

_Python for Data Analysis_ http://shop.oreilly.com/product/0636920023784.do http://shop.oreilly.com/product/0636920023784.do

geebee 10y ago
There have been some good recommendations here for machine learning and data science, but I'm not sure that's what the OP is looking for here (or maybe it is, certainly no harm in posting them). But yeah, if what you want to know how to do is query, organize, filter, trim, format, reformat, munge, and finagle data, you're probably looking for something more like the oreilly book hatmatrix mentioned above. If not the book, I'd recommend just going through pandas as much as possible. Nothing wrong with just going through the online docs. Oh, one more thing, I'm personally a huge fan of pandasql as well. It's a nice library that allows you to query a panda data frame as if it were a sql table (joins work with other data frames). Pretty much whatever is available in sqlite will be available through pandasql. There have been a few spats on the interwebs about whether it's better to do things in sql vs data frame operations. Personally, I use both - I do find some things are far easier to do with a query, and then transition over to pandas and bumpy when I get to programming/mathy things. Lastly - if you do want to do data science/ML stuff, I'd recommend going over to scikit-learn and just going through all the examples, trying things out on your own datasets.
- jbmorgado 10y ago
  OP here. You are right, my interest is actually in the Data Analysis part and what are the most up to date tools and best practices for using those tools. This is because I already have a good understanding of the scientific part (i.e. statistical analysis) from the academy and from my job as a researcher. Still, I find that a good part of the Data Science resources, starts by giving a good introduction about the Analysis part, so they are also important answers for this question. Thank you all.

geebee 10y ago

There have been some good recommendations here for machine learning and data science, but I'm not sure that's what the OP is looking for here (or maybe it is, certainly no harm in posting them). But yeah, if what you want to know how to do is query, organize, filter, trim, format, reformat, munge, and finagle data, you're probably looking for something more like the oreilly book hatmatrix mentioned above. If not the book, I'd recommend just going through pandas as much as possible. Nothing wrong with just going through the online docs. Oh, one more thing, I'm personally a huge fan of pandasql as well. It's a nice library that allows you to query a panda data frame as if it were a sql table (joins work with other data frames). Pretty much whatever is available in sqlite will be available through pandasql. There have been a few spats on the interwebs about whether it's better to do things in sql vs data frame operations. Personally, I use both - I do find some things are far easier to do with a query, and then transition over to pandas and bumpy when I get to programming/mathy things. Lastly - if you do want to do data science/ML stuff, I'd recommend going over to scikit-learn and just going through all the examples, trying things out on your own datasets.

jbmorgado 10y ago
OP here. You are right, my interest is actually in the Data Analysis part and what are the most up to date tools and best practices for using those tools. This is because I already have a good understanding of the scientific part (i.e. statistical analysis) from the academy and from my job as a researcher. Still, I find that a good part of the Data Science resources, starts by giving a good introduction about the Analysis part, so they are also important answers for this question. Thank you all.