Performance monitor node for Godot

Simple node for in-game monitoring of FPS, memory usage and such, loosely styled by Three.js's performance monitor.

It may have some performance overhead (around 180 draw calls per frame, CPU and mem usage is slight), but for FPS or memory monitoring it seems insignificant.

Can display any numeric object parameter or function call result.

Usage

• Install from AssetLib

• Add Monitor node to your scene tree

• Setup what to monitor

  •   $Monitor.os_time_per_frame()

  •   $Monitor.add_perf_monitor(Performance.TIME_FPS, "FPS")

  •   $Monitor.add_custom_monitor($Player, "hitpoints", "Player hp")
      # your $Player node must have "hitpoints" attribute, which must be either float or int

  •   var render_info_funcref: FuncRef = funcref(VisualServer, "get_render_info")
      $Monitor.add_funcref_monitor(render_info_funcref, [VisualServer.INFO_TEXTURE_MEM_USED], "Texture mem")

Also you can set plot size, color, or amount of plot stored data, check out example.gd.

Convenience function $Monitor.os_time_per_frame(), creates a monitor to show difference in OS.get_ticks_usec() between _process() calls.

Reference

•   os_time_perframe() # creates a plot panel to show difference between OS.get_ticks_usec() calls

•   add_perf_monitor() # creates a plot panel to monitor one of the built-in Performance enum 
  
    # required
    param_key: int # key from Performance enum 
    
    # optional	
    label: String = "" # label to display
    color: Color = Color(0.2, 1, 0.2, 0.5) - color
    humanise: bool = false # whether to use humanize for large numbers
    is_data_int: bool = true # !important for floats change to false
    max_value: float = 0.0 # maximum value
    data_len: int = 180 
    size_px: Vector2 = Vector2(180, 80)

•   add_custom_monitor() # creates a plot panel to monitor one of the passed objects' numeric attributes
  
    # required
    obj: Object # object to monitor
    param_name: String # name of numeric attribute of the object above
    
    # optional	
    label: String = "" # label to display
    color: Color = Color(0.2, 1, 0.2, 0.5) # color
    humanise: bool = false # whether to use humanize for large numbers
    is_data_int: bool = true # !important for floats change to false
    max_value: float = 0.0 # maximum value
    data_len: int = 180 
    size_px: Vector2 = Vector2(180, 80)

•   add_funcref_monitor() # creates a plot panel to monitor results of the calls to the passed function
  
    # required
    function_ref: FuncRef # funcref to a function that will be called every _process()
    
    # optional	
    function_params: Array # [] - array of function param values
    label: String = "" # label to display
    color: Color = Color(0.2, 1, 0.2, 0.5) - color
    humanise: bool = false # whether to use humanize for large numbers
    is_data_int: bool = true # !important for floats change to false
    max_value: float = 0.0 # maximum value
    data_len: int = 180 
    size_px: Vector2 = Vector2(180, 80)

Other notes

The draw call happens in _process(), so update rate may be not constant.

The plot uses some kind of ring buffer based on PoolIntArray or PoolRealArray for storing data.

Rendering may be optimized.