Rebecca, Yago and Octavi

your-project/Marvel_Functions.py

@@ -0,0 +1,158 @@
+import pandas as pd
+import numpy as np
+import csv
+import hashlib
+import json
+import requests
+import random
+import matplotlib.pyplot as plt
+
+def acquire_csv(csv_file):
+	data_csv = pd.read_csv('../data/'+csv_file)
+	return data_csv 
+
+def acquire_api():
+    list_of_characters = []
+    offset = 0
+    url = f'http://gateway.marvel.com/v1/public/characters'
+    private_key = "9569164e98f5f2b3a84ce161f501b6e7cda35997"
+    public_key = "f4afbc73aea9aa4703bf0ba8e236f72c"
+    limit = 100
+    ts = "1580473503"
+    # hash is a md5 digest of the ts parameter, your private key and your public key (e.g. md5(ts+privateKey+publicKey))
+    to_encode = (ts+private_key+public_key)
+    h = hashlib.md5(to_encode.encode()).hexdigest()
+
+    for i in range(15):
+        p = {
+        "apikey": public_key ,
+        "ts": ts,
+        "hash": h,
+        "limit" : limit,
+        "offset" : offset
+        }
+        offset += limit
+
+        response = requests.get(url, params=p)
+        result = response.json()
+
+        list_of_characters.append(result)
+
+    return list_of_characters
+
+def extract_names_ids(list_of_characters):
+    column_names = []
+    column_ids = []
+
+    for i in range(len(list_of_characters)):
+        for character in list_of_characters[i]['data']['results']:
+            column_names.append(character["name"])
+            column_ids.append(character["id"])
+
+    df_marvel = pd.DataFrame({'Name': column_names, 'ID': column_ids})
+    return df_marvel
+
+def wrangling(df1, stats, info):
+    df = extract_names_ids(df1)
+    cleaned_ids = df["ID"]
+    df_cleaned_ids = pd.DataFrame(cleaned_ids)
+    # We eliminated the extension of the name to avoid having repeated characters (e.g. Wolverine, Wolverine (Ultimate), Wolverine (Age of Dragon))
+    cleaned_names = df["Name"].apply(lambda x: x.split("(")[0].strip() )
+
+    # we merged both column clean name with id to create match (new dataframe)
+    df_cleaned_names = pd.DataFrame(cleaned_names)
+    df_cleaned_1 = df_cleaned_names.merge(df_cleaned_ids, left_index=True, right_index=True)
+
+    # we eliminated duplicated values (as before, we removed the extension)
+    df_cleaned_2 = df_cleaned_1.drop_duplicates("Name")
+
+    # we merge with the csv df
+    df_cleaned_3 = df_cleaned_2.merge(stats, on = "Name")
+
+    # we clean the total values with a value of 5 or less
+    Marvel_DF = df_cleaned_3[df_cleaned_3["Total"] > 5] 
+
+    # Time to work with the second csv
+    # Taking off useless columns as IDs, Alignment (repeated) Eyecolor or Race from info df
+    info_cl  = info.drop(columns = ["Publisher","ID","Alignment","EyeColor", "Race", "HairColor","SkinColor"])
+
+    # Merging dataframe with Marvel_DF
+    Marvel_temp = Marvel_DF.merge(info_cl, on = "Name")
+    Marvel_temp2 = Marvel_temp.drop_duplicates("Name")
+
+
+
+    #Converting categorical values to binomial and '-', neutral to 0
+    Marvel_characters_info_BI = Marvel_temp2.replace({"Alignment": {"good": 0 , "bad": 1, "neutral" : 0 }, 
+                                            "Gender": {"Male": 0 , "Female": 1, "-" : 0 }} )
+
+    # converting object to numeric values
+    Marvel_characters_info_BI["Gender"] = pd.to_numeric(Marvel_characters_info_BI["Gender"])
+    Marvel_characters_info_BI["Alignment"] = pd.to_numeric(Marvel_characters_info_BI["Alignment"])
+    Marvel_characters_info_BI["ID"] = Marvel_characters_info_BI["ID"].astype(str)
+
+    # Ammending some wrong values by hand
+    Marvel_characters_info_BI.loc[Marvel_characters_info_BI.Name == 'Iron Monger', 'Height'] == 300
+    Marvel_characters_info_BI.loc[Marvel_characters_info_BI.Name == 'Iron Monger', 'Weight'] == 1000
+    Marvel_characters_info_BI.loc[Marvel_characters_info_BI.Name == 'Leech', 'Height'] == 128
+    Marvel_characters_info_BI.loc[Marvel_characters_info_BI.Name == 'Leech', 'Weight'] == 30
+    Marvel_characters_info_BI.loc[Marvel_characters_info_BI.Name == 'Stardust', 'Height'] == 213
+    Marvel_characters_info_BI.loc[Marvel_characters_info_BI.Name == 'Beyonder', 'Height'] == 189
+    Marvel_characters_info_BI.loc[Marvel_characters_info_BI.Name == 'Beyonder', 'Weight'] == 109
+
+    return Marvel_characters_info_BI
+
+
+def ask_input(dict_test):
+    gender = str(input("Which gender do you want to analyse? (male/female/all) ")).lower()
+    alignment = str(input("Which alignment do you want to analyse? (good/bad/all) ")).lower()
+    if (gender not in dict_test.keys()) or (alignment not in dict_test.keys()):
+        print ("Puny human!")
+        gender, alignment = ask_input(dict_test)
+    return gender, alignment
+
+def df_selection(df, gender, alignment):
+    if alignment == 2 and gender == 2:
+        test = df
+    elif alignment == 2:
+        test = df[df.Gender == gender]
+    elif gender == 2:
+        test = df[df.Alignment == alignment]
+    else:
+        test = df[(df.Alignment == alignment) & (df.Gender == gender)]
+    return test
+
+
+def best_in(group,number = 10):
+    topX = pd.DataFrame()
+    for column in group.columns:
+        topX[column] = list(group.sort_values(by = column, ascending = False).Name.head(number))
+    top = topX.drop(topX.columns[[0,1,2,-3]], axis=1)
+    return top
+
+def mean_df(df):
+    dict_means = {}
+    col = ['Intelligence', 'Strength', 'Speed', 'Durability', 'Power', 'Combat', 'Total','Height', 'Weight']
+    for i in col:
+        dict_means[i] = df[i].mean()
+    return dict_means
+
+def median_df(df):
+    dict_medians = {}
+    col = ['Intelligence', 'Strength', 'Speed', 'Durability', 'Power', 'Combat', 'Total','Height', 'Weight']
+    for i in col:
+        dict_medians[i] = df[i].median()
+    return dict_medians
+
+def st_dev(df):
+    dict_stds = {}
+    col = ['Intelligence', 'Strength', 'Speed', 'Durability', 'Power', 'Combat', 'Total','Height', 'Weight']
+    for i in col:
+        dict_stds[i] = df[i].std()
+    return dict_stds
+
+def graphing(df, file_name):
+    # Doing a multi-graph figure
+    plt.rc('figure', titlesize=6)
+    df[df.columns[2:]].hist(ylabelsize=10, bins = 15, figsize=(20, 20))
+    plt.savefig(f"CSVs/{file_name}.png")