5 ms·

Great work, thanks! As the example app has been DoS'd, I decided to try and deploy my own on streamlit (having never used it before). The deployment process i

by Daviey 3y ago

Great work, thanks! As the example app has been DoS'd, I decided to try and deploy my own on streamlit (having never used it before). The deployment process isn't documented in your project, but it was super easy to do using the upstream docs! Welcome to try mine: https://wfcmzortl5n8cqhdowr6dm.streamlit.app/ https://wfcmzortl5n8cqhdowr6dm.streamlit.app/

GaggiX 3y ago
Well it didn't last long
- Daviey 3y ago
  Argh! It looks like a badly handled error in the app, which I can't blame the platform for... but it looks like the platform gives great access to error logging I've "rebooted" the app, but no idea how long it will last again! ``` /home/adminuser/venv/lib/python3.9/site-packages/geopandas/array.py:158: ShapelyDeprecationWarning: The array interface is deprecated and will no longer work in Shapely 2.0. Convert the '.coords' to a numpy array instead. return GeometryArray(vectorized.from_shapely(data), crs=crs) /home/adminuser/venv/lib/python3.9/site-packages/geopandas/array.py:158: ShapelyDeprecationWarning: The array interface is deprecated and will no longer work in Shapely 2.0. Convert the '.coords' to a numpy array instead. return GeometryArray(vectorized.from_shapely(data), crs=crs) [15:49:17] Streamlit server consistently failed status checks [15:49:17] Please fix the errors, push an update to the git repo, or reboot the app. ```
  - hk__2 3y ago
    > but it looks like the platform gives great access to error logging Where do you see errors in the log? In the snippet you shared, I see only a couple warning then "please fix the errors".
    - Daviey 3y ago
      As the site owner of MY deployment, it gives me a box down the bottom right which also allows me to restart the app. It's worth deploying it yourself to experiment with the platform. 1) Fork the app in Github 2) Create an account on https://streamlit.io/ 3) New app, and link to your fork in Github
llimllib 3y ago
To run it locally, I did: pip install -r requirements.txt pip install -r requirements-dev.txt pip install -r streamlit-prettymapp/requirements.txt streamlit run streamlit-prettymapp/app.py (The `streamlit` command is installed by the third requirements.txt) It does properly show the Macau map, but is hanging on `Creating map... (may take up to a minute)`; I wonder if there's an OSM api key that is not working, or something like that. edit: looks like the issue is just that it's very slow in places with a decent amount of OSM stuff? It's actually hanging on converting the OSM to geometries, as it tries to convert 400k objects edit 2: https://github.com/chrieke/prettymapp/issues/26 https://github.com/chrieke/prettymapp/issues/26
tomcam 3y ago
Also hugged to death ;)

by Daviey 3y ago

GaggiX 3y ago

Well it didn't last long

Daviey 3y ago
Argh! It looks like a badly handled error in the app, which I can't blame the platform for... but it looks like the platform gives great access to error logging I've "rebooted" the app, but no idea how long it will last again! ``` /home/adminuser/venv/lib/python3.9/site-packages/geopandas/array.py:158: ShapelyDeprecationWarning: The array interface is deprecated and will no longer work in Shapely 2.0. Convert the '.coords' to a numpy array instead. return GeometryArray(vectorized.from_shapely(data), crs=crs) /home/adminuser/venv/lib/python3.9/site-packages/geopandas/array.py:158: ShapelyDeprecationWarning: The array interface is deprecated and will no longer work in Shapely 2.0. Convert the '.coords' to a numpy array instead. return GeometryArray(vectorized.from_shapely(data), crs=crs) [15:49:17] Streamlit server consistently failed status checks [15:49:17] Please fix the errors, push an update to the git repo, or reboot the app. ```
- hk__2 3y ago
  > but it looks like the platform gives great access to error logging Where do you see errors in the log? In the snippet you shared, I see only a couple warning then "please fix the errors".
  - Daviey 3y ago
    As the site owner of MY deployment, it gives me a box down the bottom right which also allows me to restart the app. It's worth deploying it yourself to experiment with the platform. 1) Fork the app in Github 2) Create an account on https://streamlit.io/ 3) New app, and link to your fork in Github

llimllib 3y ago

To run it locally, I did: pip install -r requirements.txt pip install -r requirements-dev.txt pip install -r streamlit-prettymapp/requirements.txt streamlit run streamlit-prettymapp/app.py (The `streamlit` command is installed by the third requirements.txt) It does properly show the Macau map, but is hanging on `Creating map... (may take up to a minute)`; I wonder if there's an OSM api key that is not working, or something like that. edit: looks like the issue is just that it's very slow in places with a decent amount of OSM stuff? It's actually hanging on converting the OSM to geometries, as it tries to convert 400k objects edit 2: https://github.com/chrieke/prettymapp/issues/26 https://github.com/chrieke/prettymapp/issues/26

tomcam 3y ago

Also hugged to death ;)