I have one minor quibble in Elixir’s favour: Explorer’s interface for data frames is inspired by R’s dplyr, not Pandas.
This may seem a minor quibble, but (a) the Tidyverse is so much better-designed than the Python data analysis ecosystem, its not even funny; and (b) Python/Pandas is such a juggernaut compared to other languages’ data analysis ecosystems, that is not even funny; so when a library looks beyond Pandas for inspiration, I have two reasons to sit up and take notice.
(Pandas is fine in its way, but its core design contains one-and-a-half big flaws: row numbers are not a real index, and the index is not a real column.¹ That has knock-on effects all over the shop.)
¹ Funny parallel: Pandas has UI woes because “the index is not a real column”; Git has UI woes because “the index is not a real commit”.
A fine overview!
I have one minor quibble in Elixir’s favour: Explorer’s interface for data frames is inspired by R’s dplyr, not Pandas.
This may seem a minor quibble, but (a) the Tidyverse is so much better-designed than the Python data analysis ecosystem, its not even funny; and (b) Python/Pandas is such a juggernaut compared to other languages’ data analysis ecosystems, that is not even funny; so when a library looks beyond Pandas for inspiration, I have two reasons to sit up and take notice.
(Pandas is fine in its way, but its core design contains one-and-a-half big flaws: row numbers are not a real index, and the index is not a real column.¹ That has knock-on effects all over the shop.)
¹ Funny parallel: Pandas has UI woes because “the index is not a real column”; Git has UI woes because “the index is not a real commit”.