It is a software package with the intention to deal with data gaps in timeseries data. It is being developed with primary focus for handling data with network configurations like river networks where the data gaps can be estimated using data from connected nodes.

Nadi also means river in Nepali.

Although the term "graph" is used, here it is used to mean directional graphs as we're focusing in river networks.

nadi binary can be installed using the rust ecosystem, or the makepkg command in Arch Linux.

To compile the program, run cargo build --release, and then you'll have the nadi binary in the target/release folder. Copy that to your PATH. Also, you'll probably need shared libraries for gdal

The python plugin for QGIS is in the qgis/ directory. Copy qgis/nadi to ~/.local/share/QGIS/QGIS3/profiles/default/python/plugins/ to load it into QGIS.

A demo video of the plugin can be found here: https://youtu.be/-dmTnIKeUZs

Inputs and outputs for the subcommand connection:

Example output from the subcommand network compiled with pdflatex:

• Read graph connection from a file
  • Ignore comments
  • Read Nodes
  • Read Edges/Connections
  • Read Node Attributes from file
  • Extract node and edges from DOT language file
  • Extract node and edges attributes from DOT language file
• Visualization of the graph
  • ASCII Visualization of graph network
  • Graphviz Compatible Visualization of graph network
  • LaTeX code for Visualization of network and attributes
  • Visualization sorted by attributes (?)
  • Attributes Display using a template
• Data Filling
  • Forward Fill
  • Backward Fill
  • Center Fill
  • Linear Interpolation
  • Seasonality Fill
    • Simple seasonality
    • Seasonality Kernel (for circular averaging across data)
  • Correlation Fill
    • Simple correlation
    • Correlation Kernel (for circular averaging across data)
  • Routing Model Fill
    • Forward Propagation
    • Backward propagation
    • Both
• Data reinterpolation (e.g. to remove leap year and convert to 365 days for seasonality; to convert monthly data to bi-monthly, etc)
  • Linear interpolation
  • Median data removal
  • Nearest neighbour
  • Aggregate
• Handle Varying Time information in different nodes
• Plugin system for custom functions and methods