import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

def print_data(df):
    print("=========================================")
    print(df)
    print("-----------------------------------------")
    print(df.dtypes)
    print("-----------------------------------------")
    print(df.shape)
    print("-----------------------------------------")
    print(df.info())

def load_data(filename, columns):
    df = pd.read_csv(filename, header = None, sep = "\s+", comment ='#')
    df.columns = columns
    df = df.dropna()
    df[columns[0]] = df.year.astype(float)
    print_data(df)
    return df

df_cores       = load_data("48yrs/cores.dat",       ["year", "cores"])
df_frequency   = load_data("48yrs/frequency.dat",   ["year", "frequency"])
df_specint     = load_data("48yrs/specint.dat",     ["year", "specint"])
df_transistors = load_data("48yrs/transistors.dat", ["year", "transistors"])
df_watts       = load_data("48yrs/watts.dat",       ["year", "watts"])

violet = []
orange = []
yellow = []
green  = []
blue   = []
black  = []

violet.append((196/255, 0/255, 84/255))
orange.append((237/255, 109/255, 0/255))
yellow.append((252/255, 200/255, 0/255))
green.append((98/255, 178/255, 48/255))
blue.append((48/255, 181/255, 197/255))
black.append((25/255, 24/255, 21/255))

alpha = 0.7

plt.rcParams["axes.axisbelow"] = True
plt.rcParams["figure.figsize"] = (16,9)
plt.rcParams["lines.markersize"] = 8
plt.yscale('log')
plt.grid(True)
plt.subplots_adjust(right = 0.8)

x = df_transistors.year
y = df_transistors.transistors
plt.scatter(x, y, marker = "^", color = yellow, alpha = alpha)

x = df_specint.year
y = df_specint.specint
plt.scatter(x, y, marker = "s", color = blue, alpha = alpha)

x = df_frequency.year
y = df_frequency.frequency
plt.scatter(x, y, marker = "s", color = green, alpha = alpha)

x = df_watts.year
y = df_watts.watts
plt.scatter(x, y, marker = "v", color = violet, alpha = alpha)

x = df_cores.year
y = df_cores.cores
plt.scatter(x, y, marker = "o", color = black, alpha = alpha)

plt.xlabel("Year")

plt.legend(["Transistors (thousands)", 
            "Single-thread SPECint (thousands)", 
            "Frequency (MHz)", 
            "Typical power (Watts)", 
            "Number of logical cores"], 
           loc = "upper left",
           bbox_to_anchor=(1.0, 1.0),
           fontsize = "medium")

plt.text(0.805, 0.155, 
         "Original data up to the year 2010 collected and plotted by\nM.   Horowitz, F. Labonte, O. Shacham, K. Olukotun, \nL. Hammond, and C. Batten.", 
         fontsize = "x-small", 
         transform=plt.gcf().transFigure)

plt.text(0.805, 0.125, 
         "New plot and data collected for 2010-2017 by K. Rupp.", 
         fontsize = "x-small", 
         transform=plt.gcf().transFigure)

plt.savefig("data_visualisation.png", dpi = 300, orientation = "landscape", papertype ="a4")
plt.show()