finds the bid-ask spread for each row and records the frequency of each different size spread and graphs them at the end. Used this for liquidity research on 64 million rows of data!

bid_ask_size_spread.py

import pandas as pd
import numpy as np
import csv
import seaborn as sns
import matplotlib.pyplot as plt
import math

data = pd.read_csv("CL_June_2019.csv")
print(type(data))
#new_data = data[59528000:59800000]
#print(new_data)
#new_data.to_csv("new_file.csv")
count = 0

spread = []
frequency = {}
spread.append( data['Bid Size'])
spread.append(data['Ask Size'])
for i in range(len(spread[0])):
    count += 1
    bid = spread[0][i]
    ask = spread[1][i]
    if not math.isnan(bid) and not math.isnan(ask):
        difference = int(bid-ask)
        print(difference)
        if difference in frequency:
            frequency[difference] += 1
        else:
            frequency[difference] = 1

    print(count)

k = [keys for keys in frequency]

sns.barplot(x =  ,y = [frequency[keys] for keys in frequency])
plt.xlabel("Spread")
plt.ylabel('Frequency')
plt.title("Bid-Ask Size Spread")
plt.show()