MicroPython ESP32 color changing from temps

[deanhystad]

This defines a function:

  def color(r, g, b):
    if (ds_sensor.read_temp(rom)) > grens:
      np[0] = color((temp//2), 0, 0)  # <- Recursion is when a function call's itslef.  You don't want to use recursion here.
      np.write  # <- Missing something?
    else:
      np[0] = color(0, 0, (20-temp))  # nor should you use recursion here.
      np.write()

This calls a function:

Output:
color(0, 0, 0)

To make a function do something you need to call the function. Defining the function inside a loop does not call the function. And while Python allows defining a function inside a loop, it is convention to define functions before the start of the main body of code like this:

import machine, neopixel, onewire, ds18x20, time

def set_np(degc):
    """Convert degc temperature to rgb color.  Set neo pixels to the color"""
    color = to_be_determined
    np.set(color)
    np.write()

np = neopixel.NeoPixel(machine.Pin(5), 24)
ds_sensor = ds18x20.DS18X20(onewire.OneWire(machine.Pin(4)))
roms = ds_sensor.scan() 

while True:
    ds_sensor.convert_temp()
    time.sleep(1)
    set_np(ds_sensor.read_temp(roms[0]))

Your function for converting temperature to colors is unusual. Even if your code worked I don't think you would see the LED's light up. If the temperature is 24 C, the RGB color is (12, 0, 0). Drinking a cool soda at 12 C the RGB color is (0, 0, 5). Both of these colors are very dim.

Instead of doing something ad-hock I would use the HSV (Hue, Saturation, Value) color model. In HSV one number is devoted to color (Hue) making converting temperature-> color a simple scalar transform (multiply + add). The only strike against HSV is the conversion to RGB is complicated. However, the algorithm is easy to find online, or you could use the HSV->RGB color converter functions from the matplotlib or opencv packages.

To convert degrees C to color in the range red..blue

hsv(0, 1, 1) == rgb(255, 0, 0) and hsv(240, 1, 1) == rgb(0, 0, 255).
If I want freezing (0 C) to be blue (hue = 240) and boiling (100 C) to be red (hue = 0) the conversion is:

hue = blue + (red - blue) * (temperature - freezing) / (boiling - freezing)
hue = 240 + (0 - 240) * (temperature - 0) / (100 - 0)
hue = 240 - 2.4 * temperature

def hsv_to_rgb(hue, saturation, value):
    """Convert HSV color representation to RGB.  Could use function from matplotlib or opencv libraries."""
    chroma = value * saturation
    x = chroma * (1 - abs((hue / 60) % 2 - 1))
    v_c = value - chroma

    # Calculates rgb where r, g and b are floats with range from 0..1
    if hue <= 60 :
        rgb = (chroma + v_c, x + v_c, v_c)
    elif hue <= 120:
        rgb = (x + v_c, chroma + v_c, v_c)
    elif hue <= 180:
        rgb = (v_c, chroma + v_c, x + v_c)
    elif hue <= 240:
        rgb = (v_c, x + v_c, chroma + v_c)
    elif hue <= 300:
        rgb = (x + v_c, v_c, chroma + v_c)
    else:
        rgb = (chroma + v_c, v_c, x + v_c)

    # Convert to rgb to ints in range 0..255
    return [round(component * 255) for component in rgb]

def temp_to_rgb(degc):
    """Return RGB color value for degC"""
    # Clip temperature to range 0..100 degrees C
    degc = max(0, min(100, degc))
    # Convert degc(0..100) to hue(240..0).
    hue = 240 - degc * 2.4
    # return color as rgb.
    return hsv_to_rgb(hue, 1, 1)

for c in range(101):
    print(c, temp_to_rgb(c))