import numpy as np

import Tkinter as tk

import tkFileDialog

import pandas as pd

import matplotlib as mpl

import matplotlib.pyplot as plt

import seaborn as sns

from matplotlib.backends.backend_pdf import PdfPages

import os

#Define variables

fps = 10

#Plotting label size and rotation

xlabelsize = 6

rot = 45

#Define functions & secondary variables

def numRemover(feat):

if isinstance(feat, basestring)==1: #Temp fix to bypass nan in edited .csv files

trimmed = feat.split('(')

return trimmed[0]

#Ask user to select excel files. Store file paths in fileList

root = tk.Tk()

root.withdraw()

filePathList = tkFileDialog.askopenfilenames(parent=root, title='Choose .csv files')

#Declare lists to store info data about each file. range is the number of variables

numLarv, col, fileNameExt, fileName, x, y, v, v_avg, dst, org, bend, go, go_avg, left, left_avg, right, right_avg, coil, coil_avg = ([0]*len(filePathList) for i in range(19))

#Get directory of files

dir = os.path.dirname(filePathList[0])

#Create list of file names from the full file paths

for i in range(0,len(filePathList)):

fileNameExt[i] = os.path.basename(filePathList[i])#Extract file name w/ extension

fileName[i] = os.path.splitext(fileNameExt[i])[0] #Remove extension

# print fileName[i]

#Loop each file

for i in range(0,len(filePathList)):

raw_data = pd.read_csv(dir + '/' + fileNameExt[i])

col[i] = len(raw_data.columns)

numLarv[i] = col[i]-1

data_df = raw_data.iloc[:, :col[i]]

data_df['Feature'] = data_df.iloc[:,0].apply(numRemover) #Create feature col

data_df['Avg_group'] = data_df.mean(axis=1) #Create Group mean col

#Extract features

x[i] = data_df.loc[data_df['Feature'] == 'mom_x']

y[i] = data_df.loc[data_df['Feature'] == 'mom_y']

v[i] = data_df.loc[data_df['Feature'] == 'velocity']

v_avg[i] = v[i].mean(axis=0)

v[i]['Frame'] = range(1, len(v[i]) + 1)

v[i]['Time'] = v[i]['Frame'] / fps

dst[i] = data_df.loc[data_df['Feature'] == 'acc_dst']

org[i] = data_df.loc[data_df['Feature'] == 'dst_to_origin']

bend[i] = data_df.loc[data_df['Feature'] == 'bending']

go[i] = data_df.loc[data_df['Feature'] == 'go_phase']

go_avg[i] = go[i].mean(axis=0)

left[i] = data_df.loc[data_df['Feature'] == 'left_bended']

left_avg[i] = left[i].mean(axis=0)

right[i] = data_df.loc[data_df['Feature'] == 'right_bended']

right_avg[i] = right[i].mean(axis=0)

coil[i] = data_df.loc[data_df['Feature'] == 'is_coiled']

coil_avg[i] = coil[i].mean(axis=0)

#Start a pdf to save figures

pp = PdfPages('results.pdf')

#Trajectory Figures

for i in range(0,len(filePathList)):

plt.figure(figsize=(8, 8))

plt.plot(x[i].iloc[:, 1:col[i]], y[i].iloc[:, 1:col[i]])

plt.grid(True)

plt.ylim(0, 2048)

plt.xlim(0, 2048)

plt.title(fileName[i] + ' Trajectory')

plt.xlabel('Pixels')

plt.ylabel('Pixels')

pp.savefig()

#Velocity over time

plt.figure()

for i in range(0,len(filePathList)):

plt.plot(v[i]['Time'] , v[i]['Avg_group'])

plt.title('Velocity over time')

plt.ylabel('Pixels per frame')

plt.xlabel('Seconds')

plt.legend(fileName)

pp.savefig()

#Avg veloctiy

plt.figure()

for i in range(0,len(filePathList)):

v_avg[i].drop('Avg_group',inplace=True)

v_avgs_joined = pd.concat(v_avg, axis=1)

v_avgs_joined[:max(numLarv)].boxplot(return_type='dict')

plt.title('Velocity')

plt.ylabel('Pixels per frame')

plt.xticks(np.arange(1,len(filePathList)+1),fileName)

locs, labels = plt.xticks()

plt.setp(labels, rotation=rot)

plt.tick_params(axis='x', labelsize=xlabelsize)

pp.savefig()

#Accumulated distance

plt.figure()

for i in range(0,len(filePathList)):

plt.plot(v[i]['Time'], dst[i]['Avg_group'])

plt.title('Accumulated distance')

plt.ylabel('Pixels')

plt.xlabel('Seconds')

plt.legend(fileName, loc='upper left')

pp.savefig()

#Distance from Origin figure

plt.figure()

for i in range(0,len(filePathList)):

plt.plot(v[i]['Time'], org[i]['Avg_group'])

plt.title('Distance from origin')

plt.ylabel('Pixels')

plt.xlabel('Seconds')

plt.legend(fileName, loc='upper left')

pp.savefig()

#Bending angle

plt.figure()

for i in range(0,len(filePathList)):

bend[i]['Avg_group'].dropna(inplace=True)

print bend[i]

sns.distplot(bend[i]['Avg_group'])

plt.title('Body bend angle')

plt.ylabel('Frequency')

plt.xlabel('Degrees')

plt.legend(fileName)

pp.savefig()

#Go frequency

plt.figure()

for i in range(0,len(filePathList)):

go_avg[i].drop('Avg_group', inplace=True)

go_avgs_joined = pd.concat(go_avg, axis=1)

go_avgs_joined[:max(numLarv)].boxplot(return_type='dict')

plt.title('Go frequency')

plt.ylabel('Frequency')

plt.xticks(np.arange(1,len(filePathList)+1),fileName)

locs, labels = plt.xticks()

plt.setp(labels, rotation=rot)

plt.tick_params(axis='x', labelsize=xlabelsize)

pp.savefig()

#Left bended

plt.figure()

for i in range(0,len(filePathList)):

left_avg[i].drop('Avg_group', inplace=True)

left_avgs_joined = pd.concat(left_avg, axis=1)

left_avgs_joined[:max(numLarv)].boxplot(return_type='dict')

plt.title('Left bend frequency')

plt.ylabel('Frequency')

plt.xticks(np.arange(1,len(filePathList)+1),fileName)

locs, labels = plt.xticks()

plt.setp(labels, rotation=rot)

plt.tick_params(axis='x', labelsize=xlabelsize)

pp.savefig()

#Right bended

plt.figure()

for i in range(0,len(filePathList)):

right_avg[i].drop('Avg_group', inplace=True)

right_avgs_joined = pd.concat(right_avg, axis=1)

right_avgs_joined[:max(numLarv)].boxplot(return_type='dict')

plt.title('Right bend frequency')

plt.ylabel('Frequency')

plt.xticks(np.arange(1,len(filePathList)+1),fileName)

locs, labels = plt.xticks()

plt.setp(labels, rotation=rot)

plt.tick_params(axis='x', labelsize=xlabelsize)

pp.savefig()

#Coiling frequency

plt.figure()

for i in range(0,len(filePathList)):

coil_avg[i].drop('Avg_group', inplace=True)

coil_avgs_joined = pd.concat(coil_avg, axis=1)

coil_avgs_joined[:max(numLarv)].boxplot(return_type='dict')

plt.title(' Coil frequency')

plt.ylabel('Frequency')

plt.xticks(np.arange(1,len(filePathList)+1),fileName)

locs, labels = plt.xticks()

plt.setp(labels, rotation=rot)

plt.tick_params(axis='x', labelsize=xlabelsize)

pp.savefig()

#Finish pdf file and open

pp.close()

#On Windows, use:

os.startfile('results.pdf')

#On Mac, use:

#os.system('open ' + 'results.pdf')