pip install fcx-playground
- To use data processing steps for NAV:
from fcx_playground.fcx_dataprocess.czml_nav import NavCZMLDataProcess
obj = NavCZMLDataProcess()
data = obj.ingest("<path_to_input>")
pre_processed_data = obj.preprocess(data)
czml_str = obj.prep_visualization(pre_processed_data)
- To use data processing steps for CRS rad-range:
from fcx_playground.fcx_dataprocess.tiles_rad_range import RadRangeTilesPointCloudDataProcess
obj = RadRangeTilesPointCloudDataProcess()
data = obj.ingest("<path_to_input>")
pre_processed_data = obj.preprocess(data)
point_clouds_tileset = obj.prep_visualization(pre_processed_data)
- To visualize NAV CZML:
from fcx_playground.fcx_cesium_viz.czml_viz import CZMLViz
czml_viz_obj = CZMLViz()
nav_czml_cesium_html = czml_viz_obj.generate_html("<path_to_saved_czml>")
# use the nav_czml_cesium_html in IPython.display.HTML to render it.
- To visualize CRS rad-range 3DTiles:
from fcx_playground.fcx_cesium_viz.tiles_viz import TilesViz
tileset_viz_obj = TilesViz()
point_clouds_tileset_html = tileset_viz_obj.generate_html("<path_to_saved_point_clouds_tileset>")
# use the point_clouds_tileset_html in IPython.display.HTML to render it.
ingest
, preprocess
, prep_visualization
methods are inherited from DataProcess
Abstract Class.
As per need, we can override or write custom methods for ingest
, preprocess
, prep_visualization
, by maintaining consistency on the return type of the overrides.
- Install
python
- Install
conda
(optional but recommended) - Use either
pip
orconda
to install dependencies mentioned inrequirements.txt
- Data are ingested from AWS S3. So, Setup AWS credentials
aws configure
Preferred. This deployment configuration is assumed to be used.- Need
aws_access_key_id and aws_secret_access_key
key values; inside~/.aws/credentials
- Install Docker
- Data are ingested from AWS S3. So, Setup AWS credentials
aws configure
Preferred. This deployment configuration is assumed to be used.- Need
aws_access_key_id and aws_secret_access_key
key values; inside~/.aws/credentials
- Run
docker compose build
(will take few minutes) - Run
docker compose up
, and note down thetoken_id
- Use
localhost:8888/tree?token=<token_id>
to run Jupyter Notebook.
notebooks
dir contains all the interactive python notebooks to get started with various visualization file generations.src
dir contains modules that enables the visualization file generation.- Abstact classes defines the highlevel process on which the raw data are manupulated.
- The concrete classes are implemented from abstract classes for detailed 3d visualization file generation processes.
- There are utilities that help the visualization file generation.
- Clear Notebook
outputs
before commiting any changes to git; for clean changes tracking.