I use pyomo and gurobi to solve optimization problems. Now I have 5 variables whose names are like model.str where str can be [x, y, z, w, s]. I would like to replace str with each of these strings and evaluate it for 20 iterations. For example I need values for model.x[1], model.x[2] and etc.
I have used the following code which is not very nice but almost got my result:
dict = {}
var_names = ["x", "y", "z", "w", "s"]
for i in var_names:
for t in model.T:
var_name = "model." str(i) "[" str(t) "]"
dict[var_name] = pe.value(eval(var_name))
The result of this code is:
{'model.x1': 5000.0, 'model.x[2]': 5000.0, 'model.x[3]': 6000.0, 'model.x[4]': 7000.0, 'model.x[5]': 8000.0, 'model.x[6]': 9000.0, 'model.x[7]': 10000.0, 'model.x[8]': 11000.0, 'model.x[9]': 12000.0, 'model.x[10]': 13000.0, 'model.x[11]': 14000.0, 'model.x[12]': 15000.0, 'model.x[13]': 16000.0, 'model.x[14]': 17000.0, 'model.x[15]': 18000.0, 'model.x[16]': 19000.0, 'model.x[17]': 20000.0, 'model.x[18]': 21000.0, 'model.x[19]': 21000.0, 'model.x[20]': 21000.0, 'model.y1': 0.0, 'model.y[2]': 0.0, 'model.y[3]': 0.0, 'model.y[4]': 0.0, 'model.y[5]': 0.0, 'model.y[6]': 0.0, 'model.y[7]': 0.0, 'model.y[8]': 0.0, 'model.y[9]': 0.0, 'model.y[10]': 0.0, 'model.y[11]': 0.0, 'model.y[12]': 0.0, 'model.y[13]': 0.0, 'model.y[14]': 0.0, 'model.y[15]': 0.0, 'model.y[16]': 0.0, 'model.y[17]': 0.0, 'model.y[18]': 0.0, 'model.y[19]': 0.0, 'model.y[20]': 0.0, 'model.z1': 0.0, 'model.z[2]': 0.0, 'model.z[3]': 1000.0, 'model.z[4]': 1000.0, 'model.z[5]': 1000.0, 'model.z[6]': 1000.0, 'model.z[7]': 1000.0, 'model.z[8]': 1000.0, 'model.z[9]': 1000.0, 'model.z[10]': 1000.0, 'model.z[11]': 1000.0, 'model.z[12]': 1000.0, 'model.z[13]': 1000.0, 'model.z[14]': 1000.0, 'model.z[15]': 1000.0, 'model.z[16]': 1000.0, 'model.z[17]': 1000.0, 'model.z[18]': 1000.0, 'model.z[19]': 0.0, 'model.z[20]': 0.0, 'model.w1': 4000.0, 'model.w[2]': 5000.0, 'model.w[3]': 6000.0, 'model.w[4]': 7000.0, 'model.w[5]': 8000.0, 'model.w[6]': 9000.0, 'model.w[7]': 10000.0, 'model.w[8]': 11000.0, 'model.w[9]': 12000.0, 'model.w[10]': 13000.0, 'model.w[11]': 14000.0, 'model.w[12]': 15000.0, 'model.w[13]': 16000.0, 'model.w[14]': 17000.0, 'model.w[15]': 18000.0, 'model.w[16]': 19000.0, 'model.w[17]': 20000.0, 'model.w[18]': 21000.0, 'model.w[19]': 21000.0, 'model.w[20]': 21000.0, 'model.s1': 0.0, 'model.s[2]': 0.0, 'model.s[3]': 0.0, 'model.s[4]': 0.0, 'model.s[5]': 0.0, 'model.s[6]': 0.0, 'model.s[7]': 0.0, 'model.s[8]': 0.0, 'model.s[9]': 0.0, 'model.s[10]': 0.0, 'model.s[11]': 0.0, 'model.s[12]': 0.0, 'model.s[13]': 0.0, 'model.s[14]': 0.0, 'model.s[15]': 0.0, 'model.s[16]': 0.0, 'model.s[17]': 0.0, 'model.s[18]': 0.0, 'model.s[19]': 1000.0, 'model.s[20]': 2000.0}
However what I need is to put every model.x[i] values for a dictionary key x so that I have a dictionary with x, y, z, w, s as keys and their values as values in a list fore example to create a pandas data frame in the end.
I know for sure there is a better way of doing it, since creating variable names in every iteration is not very efficient and I am fairly new to Python. If necessary I can also provide the rest of the codes for clarity so please just let me know if I need to provide further details.
I appreciate your help in advance.
Updated Solution After receiving some great tips from dear Matthew I came up with the following solution:
var_names = ["x", "y", "z", "w", "s"]
dict = {}
for i in var_names:
var_name = f"model.{i}"
for t in model.T:
var = f"model.{i}[{t}]"
if str(var_name) in dict:
dict[str(var_name)].append(pe.value(eval(var)))
else :
dict[str(var_name)] = [pe.value(eval(var))]
import pandas as pd
df = pd.DataFrame(dict)
print(df)
The output:
model.x model.y model.z model.w model.s
0 5000.0 0.0 0.0 4000.0 0.0
1 5000.0 0.0 0.0 5000.0 0.0
2 6000.0 0.0 1000.0 6000.0 0.0
3 7000.0 0.0 1000.0 7000.0 0.0
4 8000.0 0.0 1000.0 8000.0 0.0
5 9000.0 0.0 1000.0 9000.0 0.0
6 10000.0 0.0 1000.0 10000.0 0.0
7 11000.0 0.0 1000.0 11000.0 0.0
8 12000.0 0.0 1000.0 12000.0 0.0
9 13000.0 0.0 1000.0 13000.0 0.0
10 14000.0 0.0 1000.0 14000.0 0.0
11 15000.0 0.0 1000.0 15000.0 0.0
12 16000.0 0.0 1000.0 16000.0 0.0
13 17000.0 0.0 1000.0 17000.0 0.0
14 18000.0 0.0 1000.0 18000.0 0.0
15 19000.0 0.0 1000.0 19000.0 0.0
16 20000.0 0.0 1000.0 20000.0 0.0
17 21000.0 0.0 1000.0 21000.0 0.0
18 21000.0 0.0 0.0 21000.0 1000.0
19 21000.0 0.0 0.0 21000.0 2000.0
CodePudding user response：
I had to improvise a bit without access to model.T and pe.value(eval(var_name)) but I think this is close to what you want. I used an f-string to clean up your var_name syntax. Next we check to see if a key is already in the_dict with .get() and extend or append as needed. Otherwise, we create a new key. Finally, we can just pass the_dict to the DataFrame constructor.
If this doesn't work for you, post a comment and I'll assist until it does. Providing a sample list for model.T would be a big help.
import random
import pandas as pd
the_dict = {}
var_names = ["x", "y", "z", "w", "s"]
model_t = [1, 2, 3, 1, 2, 3, 1, 2, 3]
for i in var_names:
for t in model_t:
var_name = f"model.{i}[{t}]"
if the_dict.get(var_name, False):
if isinstance(var_name, list):
the_dict[var_name].extend(random.randint(3, 9))
else:
the_dict[var_name].append(random.randint(3, 9))
else:
the_dict[var_name] = [random.randint(3, 9)]
the_dict output:
{'model.x[1]': [3, 8, 5],
'model.x[2]': [8, 3, 9],
'model.x[3]': [9, 6, 7],
'model.y[1]': [7, 4, 5],
'model.y[2]': [3, 6, 6],
'model.y[3]': [6, 7, 4],
'model.z[1]': [6, 9, 9],
'model.z[2]': [8, 8, 7],
'model.z[3]': [7, 4, 5],
'model.w[1]': [4, 9, 3],
'model.w[2]': [5, 9, 5],
'model.w[3]': [4, 9, 5],
'model.s[1]': [4, 8, 5],
'model.s[2]': [7, 6, 7],
'model.s[3]': [9, 5, 8]}
Now let's make a pandas df.
df = pd.DataFrame(the_dict)
df ouput:
model.x[1] model.x[2] model.x[3] model.y[1] model.y[2] model.y[3]...
0 3 8 9 7 3 6
1 8 3 6 4 6 7
2 5 9 7 5 6 4