data_prep_2022

[zenrabbit]

Tidy up 2022 for handoff

• test app
• report out on app and show us the data
• prep three df to match structure for cam trapping
• process a 1000
• do a quick EDA to calculate detection rate
• summarize total n pics that need to be processed and organize folders
• put organized pics onto HD for processing if app does not work?

[MarioZuliani]

1) App Tested with 60 images. Sample file uploaded to repo. 2) Ready to show and report on app (Have a mini talk to give on how to use it and how the system works). 3) Dataframes are prepped. 4) Over 1000 photos processed. 5) EDA in progress 6) Total number of Photos to process over 72,000 7) HD needs to be purchased and photos loaded

[zenrabbit]

Test of app

1. what is link to app?
2. can you test the same 1000 pics?
3. what is the detection rate of you on 1000 pics? What is the rate on the app?
4. HD - super. how big is the 70K photos

I recommend a USB 3.1 drive, SSD, 2TB https://www.amazon.ca/External-Portable-Laptop-Storage-Transfer/dp/B0BNHT3575/ref=sr_1_8?crid=1C71IZ8BYEOM1&keywords=external+hard+drive&qid=1674772507&sprefix=exterma%3B+%2Caps%2C136&sr=8-8

~$40

[zenrabbit]

actually, is you buy in USA, we can give to Nargol and she will need on too. WAY cheaper.

[zenrabbit]

this one https://www.amazon.com/SanDisk-2TB-Extreme-Portable-SDSSDE61-2T00-G25/dp/B08HN37XC1/ref=sr_1_3?crid=SD85KL2FIPY8&keywords=hd%2Bexternal%2Bhard%2Bdrive%2B2tb%2Bssd%2Bsandisk&qid=1674772764&sprefix=hd%2Bexternal%2Bhard%2Bdrive%2B2tb%2Bssd%2Bsandisk%2Caps%2C147&sr=8-3&ufe=app_do%3Aamzn1.fos.f5122f16-c3e8-4386-bf32-63e904010ad0&th=1

[zenrabbit]

1. i s this the output file?https://github.com/MarioZuliani/Shrub-Density-Effects-on-Animals/blob/main/Wildlife_images.csv
2. verdict?
3. if you do a row-by-row match, are they perfect? what percent difference between your csv and the output?
4. what is the detection rates and are on they exact same photos?
5. in your html, please add warning and message = FALSE...

[zenrabbit]

I see in html, 22.8414 and 22.39 so same rate?

• exact same photos? or same rate but you and it catch different ones
• do you have the R code written to then join the WI data to the metadata? is that easy?
• does it ID to genus? WI <- read.csv("images.csv") WI %>% group_by(genus) %>% filter(genus != "NA") %>% summarise(count=n())

[MarioZuliani]

1) Manually checked and looks like exact same photos. 2) I do not have the code written to joing WI to metadata. That may need to be manually done. 3) Yes the site should be able identify to genus if we make it.

[zenrabbit]

1. can we see the two dataframes you manually checked?
2. does 'looks like' the exact same photos means yes? lol
3. We do not want to manually join WI data to metadata. why not use R?
4. WI can identify to genus?

[zenrabbit]

see other issue