k_umr_mri/blurModel1/createGraphics.py

import pydicom
import numpy as np
import matplotlib.pyplot as plt
import os
import csv
import math

import wavelet


pos = 0
first = True
firstImage = True


labels = []
prediction = []

titles = []

with open("prediction.csv") as f:
  for row in csv.reader(f, delimiter=","):
    if first:
      first = False
      titles = row[8:12]
      continue

    if len(row) < 9:
      continue


    y = [float(row[1])]
    z = [float(row[n]) for n in [7,8,9,10,11,12]]
    #z[0] = float( min([z[0]] + [n for n in [1,2,3,4,5] if z[n] >= 0.2]))
    

    labels.append(y)
    prediction.append(z)
    pos += 1

plt.show()



n = 0
#for n in range(4):
keys = set([x[n] for x in labels])
boxes = [[] for _ in range(5)]
for y, z in zip(labels, prediction):
  k = int(y[n]) - 1
  boxes[k].append(z[n])



for y in [1.5,2.5,3.5,4.5]:
  plt.plot([1,5], [y, y], color="#cccccc")

plt.boxplot(boxes)

plt.savefig(f"graphics/prediction_{n}.pdf")
plt.close()

diffs = [abs(y[n] - z[n]) < 0.5 for y, z in zip(labels, prediction)]
good = 0
bad = 0
for x in diffs:
  if x:
    good += 1
  else:
    bad += 1

print(f"{n}: {good} <-> {bad}")


for n in [1,2,3,4,5]:
  plt.scatter([x[0] for x in labels], [x[n] for x in prediction], c=[int(x[0] + 0.5) for x in labels])
  plt.savefig(f"graphics/one_hot_{n}.pdf")
  plt.close()


for y, z in zip(labels, prediction):
  if y[0] < 4 and abs(y[0] - z[0]) >= 0.5:
    plt.plot([1,2,3,4,5], [z[n] for n in [1,2,3,4,5]])
plt.savefig(f"graphics/one_hot_all.pdf")
plt.close()


barY = []
barN = []
titles = [1.0,2.0,3.0,4.0,5.0]
for n in titles:
  yes = 0
  no = 0
  for y, d in zip(labels, diffs):
    if y[0] == n:
      if d:
        yes += 1
      else:
        no += 1
  s = yes + no
  if s > 0:
    yes = yes * 100 / s
    no = no * 100 / s
  barY.append(yes)
  barN.append(no)
width = 0.4
for y in [20, 40, 60, 80, 100]:
  plt.plot([1,5.8], [y, y], color="#cccccc")
  plt.plot([1,5.8], [y - 10, y - 10], color="#eeeeee")
plt.bar(titles, barN, width, label="Mismatch")
plt.bar([x + width for x in titles], barY, width, label="Match")
plt.legend()
plt.savefig(f"graphics/bar.pdf")
plt.close()


print(f"Yes: {barY}")
print(f"No : {barN}")