import pandas as pd
from sklearn.manifold import TSNE
import seaborn as sns
from matplotlib import pyplot as plt
import os

#Make sure working directory is the same!
os.getcwd()
os.chdir("C:\\Users\\Jonathan\\Desktop\BPS - RP1\\January\Week 12 - 11-01-2021\\15 January")

df = pd.read_csv('output_data_warrior.csv', error_bad_lines=False)
# print(df.shape)
df_select = df[["rmsd1", "rmsd2"]]
# print(df_select)

# Text below was a try to show fingerprints (still possible, but unsure what use the plot would have)
# df_length = len(df_select)
# finger = Fingerprints[5]
# print(finger)

# for x in range(df_length):
#     finger = Fingerprints[x]
#     df_select = df_select[x].update(finger)

# df_select = df_select.apply (pd.to_numeric, errors='coerce')
# df_select = df_select.dropna()
# df_select = df_select.apply (pd.to_numeric, errors='coerce')
# df_select = df_select.dropna()
# print(df_select)

#Sets out the 2 selected columns from df_select against each other
m = TSNE(learning_rate=40)
tsne_features = m.fit_transform(df_select)
tsne_features[1:4, :]

df['tsne-2d-one']=tsne_features[:,0]
df['tsne-2d-two']=tsne_features[:,1]

plt.figure(figsize=(16,10))
sns.scatterplot(
    x="tsne-2d-one", y="tsne-2d-two",
    hue="tsne-2d-two",
    palette=sns.color_palette("flare", as_cmap=True),
    data=df,
    legend="brief",
    alpha=0.7
)
print("Created plot!")

# sns.scatterplot(x="x", y="y", hue="y", palette=sns.color_palette("hls", 2), data=df, legend="full")
# plot.show()