Create a dataframe: three ways

At times it is necessary to create dataframes within a program. Learn three ways to create them.

You might be familiar with Python dictionaries and the Python Pandas library. To read this article make sure you are comfortable with the preliminary level concepts of the above mentioned topics.

Pandas is a powerful Python library for data analysis and manipulation. Dataframes are usually created by reading in datasets from disk. However, there might be cases where your program generates a lot of data which you want to capture for further manipulation or for writing onto disk.

Let us look at three ways we can create dataframes on-the-fly.

All dictionaries have two elements, the key and the value. When you want to create tables, you can do it in two ways, either by taking the keys as columns or taking the keys as rows.

Pandas dataframe from dict with keys as columns

import pandas as pd

#dictionary storing the data
data = {
    "Subjects": ["Physics","Chemistry","Maths","Csc"],
    "Marks": [87,67,90,76]
}

#dataframe from dict
df = pd.DataFrame.from_dict(data)

Output:

    Subjects  Marks
0    Physics     87
1  Chemistry     67
2      Maths     90
3        Csc     76

Pandas dataframe from dict with keys as rows

data = {
    "Physics":[87],
    "Chemistry":[67],
    "Maths":[90],
    "Csc":[76]
}
df2 = pd.DataFrame.from_dict(data,orient='index')
print(df2)

Output

            0
Physics    87
Chemistry  67
Maths      90
Csc        76

Pandas dataframe from array of arrays

In this approach each array element becomes a row. Since the rows and columns do not have a name, we can manipulate the array using row and column numbers.

#Create an array of arrays
outer_arr = [np.random.randint(0,100,size=5) for i in range(6)]

#create a dataframe out of this array
df = pd.DataFrame(outer_arr)

print(df)
#Your dataframe will be different on account of the random numbers

Output:

    0   1   2   3   4
0   9  98  24  75   8
1  63   7  25  69  95
2  49  59  89  12  10
3  84  99  86  67  61
4  73  45  53  19  95
5   5  87  84  74  35