# -*- coding: utf-8 -*- """ Created on Mon Feb 6 03:21:38 2023 @author: Albert Script that handles global variables Includes hardcoded paths, useful to share code across computers. HOW DATA PATHS WORK: There is a general project folder which includes both the data and the paths to store it. recognized_project_paths - must include the path the folder where the project will be stored FAT_CLUSTER_PATH - must be the path to the data, stored in "BigFatCluster.mat" PLACE_CELL_PATH_DOMBECK - must be path to the dombeck place cell classification matrix, "place_cell_bool_dombeck.mat" OUTPUT_PATH - folder where data will be stored. Will be created if it doesn't exist RESULTS_PATH - folder where figure panels will be stored. Will be created if it doesn't exist HOW TO SET PROJECT FOLDER IN A NEW COMPUTER: (1) Create a folder for the project and add it to "recognized_project_paths" (it's made as a list to allow it to work across multiple computers at once) (2) Copy the main data (BigFatCluster.mat) and make sure "FAT_CLUSTER_PATH" has the correct path to it (3) Copy the secondary data ("place_cell_bool_dombeck.mat") and make sure "PLACE_CELL_PATH_DOMBECK" has the correct path to it (losonczy isn't used in the final figures) """ import numpy as np import os.path import matplotlib.pyplot as plt ##### Paths ##### recognized_project_paths = [ "D:\Albert\MPI_Brain\Bonn Project", "D:\MPI_Brain\Data\Bonn Project", "C:\\Users\Albert\MyFiles\MPI_Brain\Data\Bonn Project" ] for path in recognized_project_paths: path_exists = os.path.isdir(path) if path_exists: project_path = path break else: raise FileNotFoundError('WARNING: No recognized project data path found in this computer') #Main data path FAT_CLUSTER_PATH = project_path + "\WithAmpltd\BigFatCluster.mat" #Complementary data PLACE_CELL_PATH_DOMBECK = project_path + "\place_cell_bool_dombeck.mat" PLACE_CELL_PATH_LOSONCZY = project_path + "\place_cell_bool_losonczy.mat" #Storage paths OUTPUT_PATH = project_path + "\Output\\" if not os.path.exists(OUTPUT_PATH): os.makedirs(OUTPUT_PATH) FIGURES_PATH = project_path + "\Figures\\" if not os.path.exists(OUTPUT_PATH): os.makedirs(OUTPUT_PATH) ##### FIXED EXPERIMENT SETUP VARIABLES ##### #Experiment variables MAX_POS = 1500 #Animals MOUSE_TYPE_LABELS = ["ID", "DD"] MOUSE_TYPE_INDEXES = {MOUSE_TYPE_LABELS[0]:[0,1,2,3], MOUSE_TYPE_LABELS[1]:[4,5,6,7]} MOUSE_TYPE_LABEL_BY_MOUSE = {mnum : MOUSE_TYPE_LABELS[mnum > 3] for mnum in range(8)} #Sessions # SESSION_NAMES = ['B1', 'B2', 'B3', 'T1', 'T2', 'Tn-1', 'Tn', 'P1', 'P2', 'Ex1','Ex2','B1','B2','T1','T2','P1','P2'] SESSION_NAMES = ['B1', 'B2', 'B3', 'T1', 'T2', 'T3', 'T4', 'P1', 'P2', 'Ex1','Ex2','B1','B2','T1','T2','P1','P2'] SESSION_TIMES = [0, 24, 25, 48, 49, 72, 73, 96, 97] #Hours since B1 SESSION_TYPE_LABELS = ['Baseline', 'Training', 'Probe'] SESSION_TYPE_LABELS_SHORT = ['B', 'T', 'P'] SESSION_SNUM_BY_LABEL = {SESSION_TYPE_LABELS[0]:[0,1,2], SESSION_TYPE_LABELS[1]:[3,4,5,6], SESSION_TYPE_LABELS[2]:[7,8], } SESSION_LABEL_BY_SNUM = {} #[snum : stype] for snum in range(9): for stype in SESSION_TYPE_LABELS: if snum in SESSION_SNUM_BY_LABEL[stype]: SESSION_LABEL_BY_SNUM[snum] = stype #Mouse num : list of tuples with repeatead session pairs SESSION_REPEATS = {4: [(3,5), (4,6)], 5: [(3,5), (4,6)], 6: [(1,2)]} ##### PLOTTING VARIABLES ##### #Plotting colors PCA_CMAP = 'twilight' SHUFFLE_DEFAULT_COLOR = 'gray' PREDICTION_LABELS = ['Real', 'Predicted'] PREDICTION_COLORS = {PREDICTION_LABELS[0]:'black', PREDICTION_LABELS[1]:'gray'} CCA_LABELS = ['Self', 'Unaligned', 'Aligned', 'Aligned (shift)'] CCA_COLORS = ['steelblue', 'deepskyblue', 'teal', 'darkslateblue'] ERROR_CMAP = 'viridis' SESSION_TYPE_COLORS = ['royalblue', 'indianred', 'forestgreen'] SESSION_TYPE_COLORS_DICT= {SESSION_TYPE_LABELS[i]:SESSION_TYPE_COLORS[i] for i in range(len(SESSION_TYPE_COLORS))} MOUSE_TYPE_COLORS = ['darkgreen', 'mediumseagreen'] MOUSE_TYPE_COLORS_BY_MTYPE = {MOUSE_TYPE_LABELS[0]:MOUSE_TYPE_COLORS[0], MOUSE_TYPE_LABELS[1]:MOUSE_TYPE_COLORS[1]} MOUSE_TYPE_COLORS_BY_MNUM = {mnum:MOUSE_TYPE_COLORS[MOUSE_TYPE_LABELS.index(MOUSE_TYPE_LABEL_BY_MOUSE[mnum])] for mnum in range(8)} AP_DECODING_LABELS = ['No AP', 'AP', 'Trial Shuffle', 'Session Shuffle', 'Alignment shift'] AP_DECODING_COLORS = ['slateblue', 'darkviolet', 'gray', 'darkkhaki', 'darksalmon'] DISTANCE_LABELS = ['Within', 'Between'] DISTANCE_CMAP = 'Purples' CELL_TYPE_COLORS = ['mediumslateblue', 'orchid'] CELL_TYPE_LABELS = ['Non-place cell', 'Place cell'] WEIGHT_TYPES_COLORS = ['lightseagreen', 'goldenrod'] WEIGHT_TYPES_LABELS = ['position', 'air puff'] ##### DEFAULT PARAMETERS ##### ##### Used as a reference for functions that use dictionaries as input ##### ANALYSIS_NAME_LIST = ['preprocessing', 'alignment', 'APdecoding'] #Preprocessing parameters + dimensionality reduction preprocessing_dict_name = "PCA_analysis_dict" preprocessing_param_dict = { #session params 'mouse_list':range(8), 'session_list':range(9), #Preprocessing parameters 'time_bin_size':1, 'distance_bin_size':1, 'gaussian_size':25, 'data_used':'amplitudes', 'running':True, 'eliminate_v_zeros':False, 'num_components':'all', } #CCA parameters cca_dict_name = "CCA_analysis_dict" cca_param_dict = { 'CCA_dim':6, 'return_warped_data':True, 'return_trimmed_data':False, 'sessions_to_align':'all' } #AP decoding parameter dictionary APdecoding_dict_name = "APdecoding_analysis_dict" APdecoding_param_dict = { 'limit_position':False, #If True, the regions of the belt where airpuff will be decoded are restructed 'min_pos_to_filter':1000, #lower limit (mm, MAX_POS considered the periodic maximum) 'max_pos_to_filter':1500, #higher limit (mm, MAX_POS considered the periodic maximum) ## TCA params ## 'TCA_method': "ncp_hals", #"cp_als", "mcp_als", "ncp_bcd", "ncp_hals" 'TCA_factors':12, #Dimensionality of each TCA factor 'TCA_replicates':10, #Number of times TCA is repeated 'TCA_convergence_attempts':10, #Number of times TCA can fail before giving up 'TCA_on_LDA_repetitions':2, #Number of times the whole TCA-LDA analysis is repeated, as results may vary ## LDA params ## 'LDA_imbalance_prop':.6, #If one class appears in proportion higher than this threshold, perform imbalance repetitions 'LDA_imbalance_repetitions':5, #Number of times LDA decoding is repeated using a subset of each class with equal number of samples per class 'LDA_trial_shuffles':0, #Number of trial shuffles 'LDA_session_shuffles':0, #Number of sessions shuffles # 'sessions_to_decode':[0,1,2,3,4,5,6,7,8], #Old version, take out? 'session_comparisons':'BT' #'airpuff', 'BT', 'BP', 'PT', indicates which sessions to compare against each other } default_param_dicts_names = {ANALYSIS_NAME_LIST[0]:preprocessing_dict_name, ANALYSIS_NAME_LIST[1]:cca_dict_name, ANALYSIS_NAME_LIST[2]:APdecoding_dict_name} default_param_dicts = {ANALYSIS_NAME_LIST[0]:preprocessing_param_dict, ANALYSIS_NAME_LIST[1]:cca_param_dict, ANALYSIS_NAME_LIST[2]:APdecoding_param_dict} def get_session_list_from_mouse(mnum): ''' Given a mouse number, get session numbers and labels avoiding repetition ''' slist = list(range(9)) if mnum in SESSION_REPEATS: for repeated_pair in SESSION_REPEATS[mnum]: if repeated_pair[0] in slist and repeated_pair[1] in slist: slist.remove(repeated_pair[1]) snames = [SESSION_NAMES[snum] for snum in slist] return slist, snames def get_AP_labels_from_snum_by_trial(snum_by_trial): #Define label no_ap_trials = SESSION_SNUM_BY_LABEL[SESSION_TYPE_LABELS[0]] + SESSION_SNUM_BY_LABEL[SESSION_TYPE_LABELS[2]] ap_trials = SESSION_SNUM_BY_LABEL[SESSION_TYPE_LABELS[1]] # no_ap_trials = [0,3,5] # ap_trials = [1,2,4,6,7,8] label_by_trial = [] for snum in snum_by_trial: if snum in no_ap_trials: label_by_trial.append(0) elif snum in ap_trials: label_by_trial.append(1) label_by_trial = np.array(label_by_trial) return label_by_trial