I am passing a function through a dataframe in subgroups and I would like to get a separate plot for each subgroup. Now plots overalap

Question

I want to calculate the prediction bands = (upb and lpb) of subgroups in my dataframe and plot each of them individually. I would like to plot the results in different plots because now the prediction bands overlap. I tried to save the output (lpb and upb) in a dataframe (predictionbounds) but I get an error and I think this approach is not elegant. How can I get 4 different plots out of the loop??

## 1,2,3 are functions I use in the loop
# 1) linear model
def model(x, y, start):
return  (b*x) + start

# 2)  Prediction band
def predband .....
return lpb, upb

# 3)  function to calculate CI 0f fitted parameters
def conf_int:........
return params_ci


### prepare dataset for loop 
# group dataset in subgroups. Loops is applied in each subgroup
dfforR2 = dfdata.groupby(["treatment1", "treatment2])

variables={'treatment1':treatment1, 'treatment2':'treatment2','b':float, 'start':float, 
           'r_2':float} 
results = pd.DataFrame(variables, index=[])

#Here I try to create an empty dataframe so I can save the variables 'lpb' and 'upb'
#bounds={'treatment1':treatment1, 'treatment2':'treatment2', 'lpb':float, 'upb':float}
#predictionbounds=pd.DataFrame(bounds, index=[])


### loop and make fit
for key, g in dfforR2: 

  x= np.linspace(0, 2, )
  popt, pcov = curve_fit(model, g['x'], g['y'])
  confint=(conf_int(g['y'], alpha, popt, pcov))

  lpb, upb=predband(x, g['x'], g['y'], popt, model, conf=0.95)

  new_row = {'treatment1':key[0], 'treatment2':key[1], 'slope': popt[0], 'start':popt[1], 
  'r_2':r_2} 
  results=results.append(new_row, ignore_index=True)


  Problem starts below: 
  #line below does not work as I get an error: 
  #AttributeError: 'dict' object has no attribute 'append'
  #new_bound = {'treatment1':key[0], 'treatment2':key[1], ' lpb': lpb, 'upb':upb}
  
  #this works but I would like to print different graphs
  plt.fill_between(x, lpb, upb, color = 'grey', alpha = 0.15)




 #### Plot manually the output
 # construct fitted curve for this treatment
 #x= np.linspace(0, 1500, 400)
 a = model(x, results.iloc[0,2], results.iloc[0,3])
 plt.plot(x, a, color='tab:blue', label='Ctrl_W')

Curenlty this is what I get: