Dataframe Keyerror, Although It Exists

Given the data rows = [ {'x': 1, 'y': 2, 'z': 3}, {'x': 2, 'y': 2, 'z': 3}, ] if I try constructing a dataframe like this frame = pd.DataFrame.from_records(rows, index='x'

Solution 1:

Add x to your columns:

df = pd.DataFrame.from_records(rows, index='x', columns=['x', 'y', 'z'])

Solution 2:

You need to include x in you columns. Eg:

rows = [{'x': 1, 'y': 2, 'z': 3}, {'x': 2, 'y': 2, 'z': 3}]
frame = pd.DataFrame.from_records(rows, index='x')
display(frame)
    y   z
x123223
frame = pd.DataFrame.from_records(rows, index='x', columns=['x', 'y', 'z'])
display(frame)
    y   z
x123223

Solution 3:

Use:

import pandas as pd

rows = [
    {'x': 1, 'y': 2, 'z': 3},
    {'x': 2, 'y': 2, 'z': 3},
]

frame = pd.DataFrame.from_records(rows, index='x', columns=['x', 'y', 'z'])
print(frame)

Output:

   y  z
x      
1  2  3
2  2  3

Solution 4:

columns is applied as a filter before your index argument is processed. Therefore, you see KeyError as Pandas cannot find x after filtering for ['y', 'z']. One solution is to reindex after reading data into a dataframe:

frame = pd.DataFrame.from_records(rows, index='x').reindex(columns=['y', 'z'])

Alternatively, you can specify all fields, including your index:

frame = pd.DataFrame.from_records(rows, index='x', columns=['x', 'y', 'z'])

This has the benefit of not reading in unwanted fields and filtering as a subsequent step.