Thanks @PPierre22 for reporting this issue, let me try to locate the problem. At the moment, could you try some of these checks
use_condaenv("r-reticulate")
or
make sure that your PATH Environment variable has the paths "C:/Users/.../AppData/Local/r-miniconda/envs/r-reticulate" and C:/Users/.../AppData/Local/r-miniconda/envs/r-reticulate/ {path to site packages}

Thank you @henrykironde for your prompt reply.

Check 1 : use_condaenv("r-reticulate")

FYI, I have 2 conda environments:

conda_list()
name python
1 r-miniconda C:\Users\...\AppData\Local\r-miniconda\python.exe
2 r-reticulate C:\Users\...\AppData\Local\r-miniconda\envs\r-reticulate\python.exe

Using the second one with use_condaenv("r-reticulate") do not change the problem.

Check 2 : make sure that your PATH Environment variable has the paths "C:/Users/.../AppData/Local/r-miniconda/envs/r-reticulate" and C:/Users/.../AppData/Local/r-miniconda/envs/r-reticulate/ {path to site packages}

My current paths are :

> Sys.getenv("PATH")
[1] "C:\Users\...\AppData\Local\r-miniconda\envs\r-reticulate;C:\Users\...\AppData\Local\r-miniconda\envs\r-reticulate\Scripts;C:\Users\...\AppData\Local\r-miniconda\envs\r-reticulate\Library\bin;C:\Program Files\R\R-4.0.3\bin\x64;C:\ProgramData\Oracle\Java\javapath;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem;C:\WINDOWS\System32\WindowsPowerShell\v1.0\;C:\WINDOWS\System32\OpenSSH\;C:\windows\system32;C:\windows\SysWOW64\Empirum;C:\Users\...\AppData\Local\Microsoft\WindowsApps"

I figure the second path you mention is not here, but I’m not sure how to interpret what you mean by “{path to site packages}”. In any cases, I added the following path:

C:\Users\…\AppData\Local\r-miniconda\envs\r-reticulate\Lib\site-packages;

It did not solve the problem. Also, I think that R already found this path, since each time I try to re-install DeepForest it does mention this path:

> reticulate::py_install('DeepForest', pip=TRUE) # Install the Python retriever package
Collecting DeepForest
Using cached deepforest-1.1.2-py3-none-any.whl (20.0 MB)
Collecting xmltodict
Using cached xmltodict-0.12.0-py2.py3-none-any.whl (9.2 kB)
Collecting geopandas
Using cached geopandas-0.9.0-py2.py3-none-any.whl (994 kB)
Requirement already satisfied: pandas in c:\users...\appdata\local\r-miniconda\envs\r-reticulate\lib\site-packages (from DeepForest) (1.1.5)
…

I can replicate the issue (sorry @PPierre22 we haven't been testing on Windows yet). I believe the following change will properly install on Windows (as long as RTools is installed):

install.packages('reticulate') # Install R package for interacting with Python
reticulate::install_miniconda() # Install Python
reticulate::py_install(c('gdal', 'rasterio', 'fiona')) # Install spatial dependencies via conda
reticulate::py_install('DeepForest', pip=TRUE) # Install the Python retriever package
devtools::install_github('weecology/deepforestr') # Install the R package for running the retriever

@PPierre22 - can you see if this fixes things for you?

@henrykironde - you're probably onto this already, but it's a problem with getting gdal easily installed on Windows (never fun as you know) since we can't currently get the Python deepforest package to install from conda due to those upstream version issues with pytorch on conda-forge. This approach installs all of the key spatial packages via conda before installing deepforest via pip. This doesn't appear to break any of the top-level functionality. The better long-term solution is getting a full conda based install of deepforest working.

Hi @ethanwhite,
Thanks, your trick did fix my problem! The install of DeepForest went smoothly.
I thought i was out of the wood, and was bold enough to try to install deepforestr, which returned "package ‘deepforestr’ is not available for this version of R" :D so... i'm going to update R & Rstudio, and start again...

It's not on CRAN yet, that was a typo in the original instructions. If you use the last line in the code block above you should be able to install it straight from the repository as long as you have RTools installed.

The install worked - thanks!

Awesome! We're still very much in the early development phase here so definitely let us know if you run into more issues as you.