Open christinahedges opened 4 years ago
Hi @christinahedges,
Thank you so much for your thorough review. I address the points one by one below:
Installation
1) test_output directory. I have removed the test_output file from the pip install download. It remains on the github page so that anyone who wants to test the code can download that version. The LATTE_imgs folder is repeated in the test file so that all of the test files are self contained in one folder.
Functionality
2) Data Download. That is a great point, thank you, I definitely don't want anyone to think that all the data is being downloaded. I have made this clearer on the readme page and changed the text that is printed to the command line to make it clearer that no actual TESS data is being downloaded.
3) Data Download. I have added a section to the readme file called 'LATTE workflow - what is being downloaded, when and why?' that explains the workflow and outlines when and why data is being downloaded and used and how big the files typically are (I'm hesitant to give times as they vary drastically depending on the internet speed). I also added a very short section that explains the functionalities of the four main LATTE scripts to the readme and more detailed descriptions at the start of the main .py files. Please do let me know if you had something else in mind to make it more clear.
4) Aperture Size test. I now explain all of the diagnostic plots in more detail in the readme. I have also created a document that explains how to interpret the plots in more detail with examples of 'good' and 'bad' planet candidates - this document is in 'examples' folder. I have also added short video that shows how to run LATTE.
5) GUI. I completely agree with you that the point is to not interact with the code. The default is now such that the plots are not displayed and changed the option in the GUI such that the user can choose to display by selecting "Show Plots".
Documentation
6) A statement of need. I have expanded on the blurb in the readme file to make the prupose of the code more clear.
7) Example usage. There is now a video of how LATTE is used, example output files and a document of how to interpret the results in order to identify planet candidates (in the 'example' folder).
8) Functionality documentation. Please see point 3).
9) Functionality documentation. The plots and downloads were definitely lacking explanation, I'm so sorry. I have written a short summary of each plot in the readme along with the example images. There are also more examples and more explanation of each plot in the 'example' folder.
10) Automated tests. I'm not sure why these aren't working but I'm looking into them today, sorry.
11) Other comments. I have added an annotated image of the GUI interface explaining the different functionalities.
Software paper
12) State of the field. I have added a paragraph to the paper mentioning the other tools that are available and how LATTE complements and differs from them. I haven't added very much text, because JOSS paper guidelines ask for a paper length between 250-1000 words, which I am already significantly over, but obviously do let me know if you would like the introduction to be more in depth and I'd be more than happy to add more.
This is great work Nora, thanks for the thorough response. The documentation is much more thorough now!
Hi @noraeisner
I'm completing your JOSS review, please excuse the lateness of my review! Please find some comments below. I'll add more as I keep digging!
Installation
I am going through the installation of your package. I am noticing a few things that I thought I would let you know. These don't get in the way of installing your package but might help installation, these are suggestions but by no means stop my reviewing/installing your package.
Functionality
I started running
LATTE
and went through the initial downloading set up. It wasn't very clear on what was being downloaded at this point (I believe it's essentially the manifest of target pixel files and light curve files from MAST). As I was waiting, I was a bit worried it was going to try and download all the light curves and target pixel files. It might be beneficial to add some additional documentation on whatLATTE
is downloading and why.When I ran a test target and closed the first resulting images, there was a long pause while the
LATTE
downloaded other products. I believe it downloaded the TPF. It would be useful to have somewhere in the documentation the explicit workflow forLATTE
, i.e. what is being downloaded, why, and when.Some more in depth comments about the functionality
I am not familiar with what the "Aperture Size" test is in this context, can you add in your report or documentation what it is designed to show?
I am not clear why there is a GUI for the plot window, as the point of
LATTE
seems to be to create a data validation report for TESS data. It seems like these plots should be generated and closed, rather than left up to the user to close? I see in the GUI I think this can be turned off, perhaps it should be turned off by default?Documentation
[A statement of need] OK
It is not clear from the documentation what the express purpose of
LATTE
is. I do see there is a "blurb" in the README, but that didn't make clear to me whatLATTE
was doing, (i.e. downloading data products, producing certain diagnostic plots, and producing a report). It would be beneficial for the user to add a more in-depth explanation.[Example usage] Not present
I feel the user may benefit from having an example workflow for using
LATTE
in the documentation, including the goal of the user.[Functionality documentation] OK
As I've said above, some documentation on what is being downloaded from MAST, why, typical sizes, typical times etc would improve the documentation for new users.
I find the documentation of each plot window does not explain what each diagram is supposed to show the user. I think this would be useful information to add, either in the DV report, the paper or the readme.
[Automated tests] OK
I find there are automated tests, but I can't run them, and they aren't automated through CI
Other comments
I find the documentation for the GUI a little hard to follow, could you add to the READE me with a diagram highlighting each section of the GUI and what it will do?
Software paper
[State of the field] Not present
I feel it is important in the paper to discuss the state of the field. There are other tools for looking at light curves of TESS data (e.g. lightkurve, eleanor) and there are other products for looking at the light curve quality (e.g. TESS SPOC validation reports, TESS MIT QLP reports).