AttributeError: 'float' object has no attribute 'on_change'

Community Support
, ,
1

I have 30 lines in one plot and I want to add sliders to change the trends of those lines.
I used plot.multi_line to graph out the initial plot.
I want to try CustomJS but I need to call my own function deriv1 with solve_ivp to have new y.data but I am not sure how to do so.
Then I tried the update_data function but it results an error AttributeError: ‘float’ object has no attribute ‘on_change’.
I am not sure what to do so that I can have my plot to change with the sliders.

import numpy as np

from bokeh.io import curdoc
from bokeh.layouts import row, column
from bokeh.models import ColumnDataSource
from bokeh.models.widgets import Slider, TextInput, Panel, Tabs
from bokeh.plotting import figure
import math
from scipy.integrate import solve_ivp, odeint
from bokeh.models import CustomJS, Callback, HoverTool, Button


# To do list:


# --------------------- Static Parameters    --------------------- #

b0 = 93 * (10**-5)  # 93      unit : 1/bars
deltH_0 = 95300  #               unit: j/mol 
Tw = 293.0 # water temperature, utility
T_in = 298.0  #   ambient temperature,  also inlet temperature, in kelvin  unit: kelvin, depends on location
T0 = 353.15  # reference temeperature to be used in the Toth isotherm   unit: kelvin
t_h0 = .37  # heterogeneity constant 
apha = 0.33
chi = 0.0
q_s0 = 3.40 # qs_var = q_s0 = 3.4 due to chi = 0 mol/kg
# R = 8.314* (10**3) # Universal gas constant - LPa/molK
Rg = .0821 # Universal gas constant in l-atm/mol/K
kT0 = 3.5 * (10 ** -2)  # used to compute kT for r_CO2... in mol/Kg-pa-sec
EaCO2 = 15200 # activation for KT -- J/mol
ps = 880.0 #
deltH_co2 = 75000.0  # calculate temeprature change   unit: jol/mol

R_constant = 8.314 # jol/kelvin-mol = m^3-pa/mol-kelvin

# ------------------ For Equation : Enegergy Ballance  -------------- #
pg = 1.87  # 
h = 13.8 # 
Cp_g = 846.0  # J/kgKelvin
Cp_s = 1500.0  # J/kgKelvin

def cube(x):
    if 0<=x: return x**(1./3.)
    return -(-x)**(1./3.)
# ------------------ ODE Repetitive Shortcut -------------- #


# ------------------ Equastions to calculate Rco2 -------------- #
def b(T): # will be called inside deriv
    b = b0 *(math.exp(((deltH_0 / (R_constant * T0)) * (T0 / T - 1))))
    return b

def t_h(T): # will be call inside  deriv
    t_h = t_h0 + (apha * (1 - (T0 / T)) )
    return (t_h)

# Calculate rco2_n (not ode)
def R_co2(T, c_co2, q):
    kn = kT0 * ( math.exp(( (-EaCO2) / (R_constant*T) )))
    b_var = b(T)
    t_var = t_h(T)
    rco2_1 = R_constant * T * c_co2 
    rco2_2 = ((1 - ((q / q_s0) ** (t_var))) ** (1 / t_var))
    rco2_3 = q / (b_var * q_s0)
    rco2 = kn * (rco2_1 * rco2_2 - rco2_3) 
    return rco2

"""
    Defines the differential equations for odes of DAC
    Arguments:
        y :  vector of the state variables:
                  y = all 15 states
        t :  time
        params :  vector of the parameters:
                  params = [V, r, T, c_co2_0, episl_r, v0]
    
"""

def deriv1(t, y, params):
    T_n, co2_n, q_n, T_n2, co2_n2, q_n2,T_n3, co2_n3, q_n3, T_n4, co2_n4, q_n4,T_n5, co2_n5, q_n5 = y # the rest of 12 vars a_n are not used, only for the success of solve_ivp
    V, T, c_co2_0, episl_r, volumetric_flow = params

    ###############   -----  Parameters depend on input  -----  ###############
    LoverR = 2.5/20 # Straight from the paper - shallow bed, so L << R
    r = cube(V/(LoverR*math.pi))
    v0 = volumetric_flow / (math.pi *r*r )
    L = V / (math.pi * (r ** 2))
    deltZ = L / 5.0  # 5 boxes in total
    a_s = 150 #Straight from the paper
    theta = (1 - episl_r) * ps * Cp_s + episl_r * pg * Cp_g
    
    temperature_constant = ((v0  * pg* Cp_g) / (theta * deltZ))
    temperature_constant2 = (1 - episl_r) * ps * deltH_co2 /theta 
    temperature_constant3 = a_s * h /theta
    concentration_constant = v0 / (episl_r * deltZ)
    concentration_constant2 = (1 - episl_r) * ps/episl_r 
    
    T1dot = -temperature_constant* T_n + temperature_constant* T_in + temperature_constant2* (
        R_co2(T_n, co2_n, q_n))+ temperature_constant3*(Tw - T_n)
    # print(f"T1", {T1})
    co2_1dot = -concentration_constant * co2_n + concentration_constant * c_co2_0 - (
        R_co2(T_n, co2_n, q_n)) * concentration_constant2
    q1dot = R_co2(T_n, co2_n, q_n)
    # print(f"energy balance in T1", {ener_balan(v0, theta, deltZ)})
    
    T2dot = -temperature_constant * T_n2 + temperature_constant * T_n +temperature_constant2 *(
        R_co2(T_n2, co2_n2, q_n2)) + temperature_constant3*(Tw - T_n2)
    # print(f"T2", {T2})
    co2_2dot = -concentration_constant* co2_n2 + concentration_constant * co2_n - (
        R_co2(T_n2, co2_n2, q_n2)) * concentration_constant2
    q2dot = R_co2(T_n2, co2_n2, q_n2)
    # print(f"energy balance in T1", {ener_balan(v0, theta, deltZ)})

    T3dot = -temperature_constant* T_n3 + temperature_constant * T_n2 + temperature_constant2* (
        R_co2(T_n3, co2_n3, q_n3)) + temperature_constant3 * (Tw - T_n3)
    co2_3dot = -concentration_constant * co2_n3 + concentration_constant * co2_n2 - (
        R_co2(T_n3, co2_n3, q_n3)) * concentration_constant2
    q3dot = R_co2(T_n3, co2_n3, q_n3)

    T4dot = -temperature_constant* T_n4 + temperature_constant* T_n3 + temperature_constant2*(
        R_co2(T_n4, co2_n4, q_n4)) + temperature_constant3 * (Tw -  T_n4)
    co2_4dot = -v0 / (episl_r * deltZ)* co2_n4 + v0 / (episl_r * deltZ)* co2_n3 - (
        R_co2(T_n4, co2_n4, q_n4)) * concentration_constant2
    q4dot = R_co2(T_n4, co2_n4, q_n4)

    T5dot = -temperature_constant * T_n5 + temperature_constant * T_n4 + temperature_constant2 * (
        R_co2(T_n5, co2_n5, q_n5))+ temperature_constant3*(Tw - T_n5)
    co2_5dot = -concentration_constant * co2_n5 + concentration_constant * co2_n4 - (
        R_co2(T_n5, co2_n5, q_n5)) * concentration_constant2
    q5dot = R_co2(T_n5, co2_n5, q_n5)

    # result = np.array([T1, T2, T3, T4, T5, co2_1dot, co2_2dot, co2_3dot, co2_4dot, co2_5dot, q1dot, q2dot, q3dot, q4dot, q5dot]).reshape(-1, 1)

    return [T1dot, co2_1dot, q1dot, T2dot, co2_2dot, q2dot, T3dot, co2_3dot, q3dot, T4dot, co2_4dot, q4dot, T5dot, co2_5dot, q5dot]

# ------------------ User generated - Slider initial value -------------- #
V = .003  # volume
T = 298.0 # +273 ambrient temperature
c_co2_0 = .016349 # mol/m^3    
episl_r = 0.3  # void
volumetric_flow = .01 # m^3/s

# air humidity 
# no radius and length, nonly nr *reed V, july 6th

# ------------------ Initial Conditions to set up solve_ivp -------------- #
t0, tf = 0.0, 21600.0 # 6hrs
co2_initial = 0
q_init_cond = 0
init_cond = [T, co2_initial, q_init_cond, T, co2_initial,q_init_cond, T, co2_initial, q_init_cond, T, co2_initial, q_init_cond, T,co2_initial, q_init_cond]
# ,20.000, 0.000, 0.000,20.000, 0.000, 0.000,20.000, 0.000, 0.000,20.000, 0.000, 0.000
params = [V, T, c_co2_0, episl_r, volumetric_flow]
N = 30 # Number of points 
tspan = np.linspace(t0, tf, N)

soln = solve_ivp(deriv1, (t0, tf), init_cond, args=(params,), t_eval = tspan, method = "BDF", rtol = 1e-5, atol = 1e-8)  # init_cond = (T, c_co2_0, q0)
# soln = solve_ivp(deriv1, (t0, tf), init_cond, args=(params,), method = "BDF", rtol = 1e-5, atol = 1e-8)  # init_cond = (T, c_co2_0, q0)
# deriv1([t0, tf], )
# print(soln)

## --------------------  Extract Figures from returned solve results and match them with Z 
def Extract(lst, term):
    return [item[term] for item in lst]
dotT= [soln.y[0], soln.y[3], soln.y[6], soln.y[9], soln.y[12]]
dotCo2 = [soln.y[1], soln.y[4], soln.y[7], soln.y[10], soln.y[13]]
dotQ = [soln.y[2], soln.y[5], soln.y[8], soln.y[11], soln.y[14]]

temp = {}
co2 = {}
q = {}
for i in range(0, N):
    temp['temp'+str(i)] = Extract(dotT, i)
    co2['co2'+str(i)] = Extract(dotCo2, i)
    q['q'+str(i)] = Extract(dotQ, i)

temp_list = list(temp.values()) # make it as np array
co2_list = list(co2.values())
q_list = list(q.values())
for i in range(0, N):
    temp_list[i].insert(0, T)
    co2_list[i].insert(0, co2_initial)
    q_list[i].insert(0, q_init_cond)

r = cube(V/(20*math.pi))
L = V / (math.pi * (r ** 2))
vec_Z = np.linspace(0, L, 6)

np.array(co2_list)
np.array(q_list)
np.array(temp_list)

TOOLS = "pan,undo,redo,reset,save,wheel_zoom,box_zoom"
source_x = ColumnDataSource(data=dict(vec_Z=vec_Z))
source_y = ColumnDataSource(data=dict(temp_list=temp_list))
p_temp = figure(width=500, height=500, x_range=[0, L],  y_range=[296, 299], tools=TOOLS,)
p_temp.multi_line(
    xs=[vec_Z]*N,
    ys=temp_list,
    color=['orangered',  'seagreen','deepskyblue', 'indigo', 'gold']*6
)

def mapWithL(input_array):
    temp = {}
    for i in range(0, N):
        temp['temp'+str(i)] = Extract(input_array, i)

    temp_list = list(temp.values()) # make it as np array
    for i in range(0, N):
        temp_list[i].insert(0, T)

    np.array(temp_list)
    return temp_list

# Set up widgets
# text = TextInput(title="title", value='my sine wave')
# offset = Slider(title="offset", value=0.0, start=-5.0, end=5.0, step=0.1)
# amplitude = Slider(title="amplitude", value=1.0, start=-5.0, end=5.0)
# phase = Slider(title="phase", value=0.0, start=0.0, end=2*np.pi)
# freq = Slider(title="frequency", value=1.0, start=0.1, end=5.1)

# Set up sliders 
V_slider = Slider(title="Volume of bed"+" (initial: "+str(V)+")", value=V, start=.001, end=.005, step=.001)
T_slider = Slider(title="Ambient temperature"+" (initial: "+str(T)+")", value=T, start=293, end=310, step=1)
c_co2_0_slider = Slider(title="Initial CO2 concentration"+" (initial: "+str(c_co2_0)+")", value=c_co2_0, start=0.016, end=0.025, step=0.02)
episl_r_slider = Slider(title="Episl r"+" (initial: "+str(episl_r)+")", value=episl_r, start= .3, end= .5, step=.03)
volumetric_flow_slider = Slider(title="Initial flow"+" (initial: "+str(volumetric_flow)+")", value=volumetric_flow, start=1, end=5, step=1)

# Set up callbacks
# def update_title(attrname, old, new):
#     plot.title.text = text.value

# text.on_change('value', update_title)

# Set up plot


def getVecZ():
    V0 = V_slider.value
    r = cube(V0/(20*math.pi))
    L = V0 / (math.pi * (r ** 2))
    vec_Z = np.linspace(0, L, 6)
    return vec_Z

def update_data(attrname, old, new):

    # Get the current slider values
    V0 = V_slider.value
    T_temp = T_slider.value
    c_co2_temp = c_co2_0_slider.value
    volumetric_flow_temp = volumetric_flow_slider.value
    episl_temp = episl_r_slider.value
    params_temp = [V0, T_temp, c_co2_temp, episl_temp, volumetric_flow_temp]

    soln_temp = solve_ivp(deriv1, (t0, tf), init_cond, args=(params_temp,), t_eval = tspan, method = "BDF", rtol = 1e-5, atol = 1e-8)  # init_cond = (T, c_co2_0, q0) 
    # Generate the new curve
    T_res_temp= [soln_temp.y[0], soln_temp.y[3], soln_temp.y[6], soln_temp.y[9], soln_temp.y[12]]
    temp_list_temp = mapWithL(T_res_temp) # y
    r = cube(V0/(20*math.pi))
    L = V0 / (math.pi * (r ** 2))
    vec_Z_temp = np.linspace(0, L, 6) # x
    # vec_Z = getVecZ()
    source_x.data = dict(vec_Z=vec_Z_temp)
    source_y.data = dict(temp_list=temp_list_temp)

for w in [V_slider, T_slider, c_co2_0_slider, episl_r_slider, volumetric_flow_slider]:
    w.on_change('value', update_data)

inputs_reaction = column(V_slider , T_slider, c_co2_0_slider, episl_r_slider, volumetric_flow_slider)
# inputs = column(text, offset, amplitude, phase, freq)

curdoc().add_root(row(inputs_reaction, p_temp, width=800))
curdoc().title = "Sliders"
2

At the very bottom, you are trying to call on_change on volumetric_flow, which is a number, not a widget:

for w in [V_slider, ..., volumetric_flow]:

Perhaps you meant to use volumetric_flow_slider.

FYI in the future, please always include a complete stack trace.

4

Sorry for my typo. I corrected it but no changes are made even though the server is successfully launched and there is no error in console.

5

You don’t appear to actually be adding anything to curdoc anywhere? Are you running this as a Bokeh server app, i.e. by invoking bokeh serve --show myapp.py to run it? It seems like you need to take a step back and walk through and understand some of the simpler examples in the Bokeh server documentation:

6

There are also lots of complete examples you can study and emulate here:

But in any case, the best advice I can give you is to start much smaller and simpler and work your way up to your actual use-case.

7

Sorry I had commented out dozens of lines so I missed the last few lines

8

Just updated the code. I follow the structure of sliders.py in the example folder.

9

@siccccc if you want me to take a look further, you need to pare things down and simplify, a lot. Specifically, we need a Minimal Reproducible Example, with emphasis on the “minimal”. Often times the very act of creating an MRE can lead you to find the problem/solution yourself. But even if not, the key to both learning and to asking good questions is to start smaller, and simpler. [1]

  1. Dumping 300 lines of code and saying “it doesn’t work” is not a reasonable ask for an all-volunteer support forum. ↩︎

10

pass it as ColumnDataSource like sel=ColumnDataSource()
then in your selection js callback put sel.data={‘i’:source.selected.indices}, now you can access your updated sel in python.

11

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.