Process data in a different, more concise format with survival observations over time

[ggilestro]

My data observation have the following format

| Drug | hour | alive |
| aba | 0 | 71 |
| aba | 5 | 70 |
| aba | 10 | 68 |
| aba | 15 | 65 |
| aba | 20 | 60 |

I have 16 drugs, 3 concentrations for each drug and about 70 individuals per condition.
Changing the file format to the default input would produce 5k-10k lines.

Is there any way to either read the CSV in the format I have or convert it into something that cab be understood by kfm?

Thanks

[CamDavidsonPilon]

Hi @ggilestro

So your csv has ~3k rows already (16 * 3 * 70)? Is there a problem with producing 5k-10k lines?

[ggilestro]

No problem, just a way to say that I cannot possibly do this manually.
Anyway, other people have pointed me to the tinyr package in R which has a python equivalent in tidypython or one could follow these hints.
Those would allow me to do what I mentioned.

[CamDavidsonPilon]

ooo is "alive" a count? This might be also called a "flatmap".

If you can make it look something like:

| Drug | hour | deaths |
| aba | 0 | 0 |
| aba | 5 | 1 |
| aba | 10 | 2 |
| aba | 15 | 3 |
| aba | 20 | 5 |
...

that is probably easier to use with lifelines. Then the "deaths" is a weight column for any analysis. There's no censoring?