Cloud Price Analysis

import pandas as pd
import matplotlib.pyplot as plt
from collections import defaultdict

df = pd.read_json("../../data/raw/index.json")

features = {
    "formatVersion",
    "disclaimer",
    "offerCode",
    "version",
    "publicationDate",
    "products",
    "terms",
}
df.head()

print("Shape of the data\n\n", df.shape)
print("\n\n")
print("Is null?\n\n", df.isnull().sum())
print("\n\n")
print("Columnwise Datatype\n\n", df.dtypes)

df_products = df["products"]
df_products = df_products.fillna("NA")

dfp = df_products.sample(
    2400
)  # make a subset of 2400 entries for easier and smaller calculations

dfp_normalize = pd.json_normalize(dfp)
(dfp_normalize).fillna("NA")

pd.Series((dfp_normalize["productFamily"])).unique()

list_k = []
for k, v in dfp.items():
    attributes = v["attributes"]
    for k, v in attributes.items():
        list_k.append(k)
unique_keys = set(list_k)
for i in unique_keys:
    print(i)

def make_attributes_dictionary(dfp):
    new_dict = {}
    for i in range(len(dfp)):
        new_dict.update(dfp[i]["attributes"])
    return new_dict

def attributes(attribute):
    list_v = []
    for key, value in dfp.items():
        attributes = value["attributes"]
        for k, v in attributes.items():
            if k == attribute:
                list_v.append(v)
    print(attribute, ":")
    print(set(list_v))

pd.Series(dfp_normalize["attributes.location"]).unique()

pd.Series(dfp_normalize["attributes.locationType"]).unique()

pd.Series(dfp_normalize["attributes.currentGeneration"]).unique()

pd.Series(dfp_normalize["attributes.instanceFamily"]).unique()

pd.Series(dfp_normalize["attributes.vcpu"]).unique()

pd.Series(dfp_normalize["attributes.physicalProcessor"]).unique()

pd.Series(dfp_normalize["attributes.clockSpeed"]).unique()

pd.Series(dfp_normalize["attributes.memory"]).unique()

pd.Series(dfp_normalize["attributes.operation"]).unique()

pd.Series(dfp_normalize["attributes.ecu"]).unique()

numerical_features = [
    "vcpu",
    "gpu",
    "ecu",
    "normalizationFactor",
    "instanceCapacity2xlarge",
    "normalizationSizeFactor",
    "instanceCapacity4xlarge",
    "instanceCapacityXlarge",
    "instanceCapacityLarge",
]

def common_attribute_count(attribute_key, attribute_value):
    counter = 0
    for key, value in dfp.items():
        attributes = value["attributes"]
        new_dict = defaultdict(list)
        for k, v in attributes.items():
            if k == attribute_key:
                if v == attribute_value:
                    new_dict[k].append(v)
                    counter += 1
    #     print(dict(new_dict), counter)
    return counter

def attributes_plots(attribute):
    list_v = []
    for key, value in dfp.items():
        attributes = value["attributes"]
        for k, v in attributes.items():
            if k == attribute:
                list_v.append(v)
    unique_values = set(list_v)
    return unique_values

def bar_chart(attribute_key):
    a = attributes_plots(attribute_key)
    if attribute_key in numerical_features:
        a.discard("NA")
        a.discard("Variable")
        print((a))
        a = sorted(a, key=lambda x: float(x))
    else:
        a = sorted(a)

    plt.subplots(figsize=(20, 10))
    for i in a:
        if i == "":
            i = "NA"
        else:
            y = common_attribute_count(attribute_key, i)
            plt.bar(i, y)
            plt.xticks(rotation=90)
            plt.title(attribute_key)
    plt.show()

for i in unique_keys:
    bar_chart(i)

attribute_key = "location"
attribute_value = "Asia Pacific (Tokyo)"
common_attribute_count(attribute_key, attribute_value)

attribute_key = "instanceFamily"
attribute_value = "General purpose"
common_attribute_count(attribute_key, attribute_value)

def common_attribute_finder(attribute_key, attribute_value):
    filename = attribute_key + "_" + attribute_value + ".txt"
    with open(filename, "w+") as f:
        for key, value in dfp.items():
            attributes = value["attributes"]
            new_dict = defaultdict(list)
            for k, v in attributes.items():
                if k == attribute_key:
                    if v == attribute_value:
                        new_dict[k].append(attributes)
                        print((new_dict), file=f)
        f.close()

def find_second_common_attribute(attribute_key, attribute_value, attribute_two):
    counter = 0
    filename = attribute_key + "_" + attribute_value + ".txt"
    with open(filename, "r") as file:
        d = file.read()
    d = d.split("\n")
    for i in d:
        if attribute_two in i:
            #             print(i)
            counter = counter + 1
    print("Total values that have the second attribute common : ", counter)
    return counter

attribute_key = "memory"
attribute_value = "32 GiB"
common_attribute_finder(attribute_key, attribute_value)
common_attribute_count(attribute_key, attribute_value)
find_second_common_attribute(attribute_key, attribute_value, "EBS")