Repository navigation
Round-5 review questions for Adisankar (pandas): two questions, and three small fixes to confirm #57
Description
@adisankarmt — these are the open questions from your round-5 review of pandas (#26), collected here so they outlive that PR's merge. All 85 of your suggestions are already applied on the PR exactly as you wrote them. Nothing here changes the page — your answers decide what the translation tool learns next. No deadline, and a short reply is all we need: one word per row is plenty.
What your review told us. You changed 85 of the 125 Malayalam lines, more than in numpy, as you said in your summary. Everything we taught the tool after round 4 came through clean: full stops and colons, capitals, For example, the English plurals, provide → നൽകുന്നു, the eleven commas, colons and one sentence split that a script made, all of which you kept. What you corrected this time is mostly two things: ordinary English words the tool had turned into Malayalam (efficient, convenient, satisfy, compare, success — you put every one back in English), and long sentences joined with a comma, which you split into two almost every time. We will teach the tool both, with word lists and small scripts checked against your own sentences first.
1. Two questions.
| # | Item | One way | The other way | Our guess |
|---|---|---|---|---|
| 1 | Lines 68–72, the five short bullets under endows them with methods that facilitate operations such as (reading in data, adjusting indices, …), and line 326, the note lead-in For the .apply() method: — you returned all of them to the English exactly |
A general rule: a short bullet or lead-in like this, a label rather than a sentence, stays in English, as exercises and Further Reading do | Only here, because the tool's Malayalam (data read ചെയ്യൽ) read badly and English was the quickest fix; another lecture's bullets could still be Malayalam |
A general rule, but it is your call. In matplotlib you turned output in all the usual formats into a Malayalam sentence, so we want to be sure |
| 2 | A comma between two items that carry a case ending as well as -ഉം: line 474 `DataFrame`-ന്റെയും, `Series` objects-ന്റെയും and line 615 US-ന്റെയും, Australia-യുടെയും (you added both). In numpy you left two such pairs without a comma (lines 840 and 902) |
Always a comma, as for Columns-ഉം, rows-ഉം |
Optional | Always, now. If so, we will add the comma in numpy at line 902 in a small follow-up PR for you to approve |
2. Three small fixes we made to your text — please confirm. They are in a separate commit (5da1c5c). Say if any is wrong and we will put yours back.
| Line | You wrote | We changed it to | Why |
|---|---|---|---|
| 54 | … ഒരു package ആണ് [Pandas](https://pandas.pydata.org/) . |
… ആണ് [Pandas](https://pandas.pydata.org/). |
A space had slipped in before the full stop |
| 161 | … analyze ചെയ്യാനും ഇത് ഒരു powerful tool ആണ്. |
one space | A double space |
| 340 | … condition satisy ചെയ്യുന്ന rows-ന്റെ … |
satisfy |
A typo |
Everything else is exactly as you wrote it, including Excel, JSON, Parquet with capitals at line 597 and computer for the English machine at line 538.
Thank you — the words you keep changing are now a short list the tool can be checked against, and that is what makes the next lecture better.
Activity
adisankarmt commentedon Oct 20, 2026
1.1 - the points were barely 2-3 words. so i retained it in english. to add on, the translated part read badly. so yes, make it general rule - reading small sentences in english is better than bad translation
1.2 - always a comma
mmcky commentedon Oct 20, 2026
Thank you, @adisankarmt — all three are clear, so this issue can close.
matplotlib, still become Malayalam sentences.-ഉം, as forColumns-ഉം, rows-ഉം.We will teach the tool both rules before the next lecture is drafted. One small pull request for you to approve, #59, brings
numpyinto line: line 840 (`a`-യിലും, `b`-യിലും) and line 902 (speed-ന്റെയും, memory-യുടെയും). On theintropage you are reviewing now (#58), answer 1.2 adds one comma, at line 18 (economics-ലും, finance-ലും); the pull request's description says so.Apply the editor's ml#57 answer to numpy: commas on case-marked -ഉം p…