CODEWITHSUNDEEP
pythonpython-3.xpandasdictionarydataframe
novawaly
novawaly

Reputation: 1251

Adding a Dictionary as a new row in a data frame

I have a function that returns a dictionary of keys and results.

I'd like to create a new function that loops through different values. Each value would produce a new dictionary of different result but with the same keys.

I'd like to have this function create a dataframe and with each iteration through the loop, the index (or first column) is set to the i value of my loop and the row would be the resulting dictionary. .

the dictionary would look like {key1: 46, key2:100,key3:200}

start = 10
stop = 100
step = 10

the final result would look something like:

    key1  key2  key3
10   46   100    200
20   50    75     60
30   80     2     10
40   100    50     6
50   10     8      33
etc...

Upvotes: 6

Views: 8434

Answers (2)

ALollz
ALollz

Reputation: 59519

Create a nested dictionary of the value and the dictionary the function returns, and then use the DataFrame constructor from_dict with orient='index' at the very end.

import pandas as pd
import numpy as np
np.random.seed(123)

start = 10
stop = 100
step = 10

d2 = {}
while start <= stop:
    # Simulating your function that returns a dictionary:
    d = {'key1': np.random.randint(1,100), 
         'key2': np.random.randint(1,10), 
         'key3': np.random.randint(20,70)}
    # Add that dictionary to a dictionary
    d2[start] = d
    start+=step

pd.DataFrame.from_dict(d2, orient='index')

Output:

     key1  key2  key3
10     67     3    58
20     18     4    62
30     58     7    53
40     97     2    67
50     74     1    66
60     97     4    34
70     37     1    36
80     69     2    23
90      3     5    59
100    67     5    67

Upvotes: 4

Sunny Liu
Sunny Liu

Reputation: 79

Not sure exactly your rules of selecting objects from the list of values.. so I just made a random selection here. But the idea should be similar:

value_list = np.random.rand(1000).tolist()
df = pd.DataFrame()
for i in range(10, 101, 10):
    a, b, c = random.sample(value_list, 3)
    df.loc[i, 'key1'] = a
    df.loc[i, 'key2'] = b
    df.loc[i, 'key3'] = c

Upvotes: 0

Related Questions