In [72]:

    

import os
import sys
# import webbrowser
import shutil
import pandas as pd
# import pdb


    

In [73]:

    

%matplotlib inline


    

In [74]:

    

from matplotlib import pyplot


    

In [75]:

    

#LIBPATH = os.path.abspath("../..")
scriptdir = os.path.abspath(".")
examples_dir = os.path.dirname(scriptdir)
LIBPATH = os.path.dirname(examples_dir)
print("LIBDIR:", LIBDIR)
print("Scriptdir:", scriptdir)
#sys.path.insert(0, LIBPATH)


    

    




LIBDIR: /Users/rasmus/Dev/nascent
Scriptdir: /Users/rasmus/Dev/nascent/examples/single_duplex

In [76]:

    

from nascent.stat_analysis.processing import get_datafile


    

In [77]:

    

structure = "circfb_1"
structure = "fourway_junction_1"

runidxs = [-1] #
# runidxs = [-2] #

statsfiles, statsfolders = zip(*[
    get_datafile(statsfile="monitored_strands_stats.tsv", runidx=runidx,
                 basedir=scriptdir, structure=structure)
    for runidx in runidxs])


    

    




Using data root directory: /Users/rasmus/Dev/nascent/examples/single_duplex/simdata/fourway_junction_1
Using stats in folder: /Users/rasmus/Dev/nascent/examples/single_duplex/simdata/fourway_junction_1/2016-03-14 193519

In [78]:

    

print(statsfiles)
stats = [pd.read_table(statsfile) for statsfile in statsfiles] # http://pandas.pydata.org/pandas-docs/stable/io.html
runstats = stats[0]


    

    




('/Users/rasmus/Dev/nascent/examples/single_duplex/simdata/fourway_junction_1/2016-03-14 193519/monitored_strands_stats.tsv',)

In [79]:

    

runstats['system_time_end'] = runstats['system_time'].shift(-1).dropna()
data_by_strand = runstats.groupby(['strand_uid'], sort=False)
strand, s_trace = next(iter(data_by_strand))
s_state_traces = s_trace.groupby(['complex_state'], sort=False)


    

In [80]:

    

for cstate, s_state in s_state_traces:
    #     s_state['s_state_time_cum'] = s_state['tau'].cumsum()
    #     s_state['s_state_partition'] = s_state['s_state_time_cum']/s_state['system_time_end']
    # Edit: Avoid using chained indexing:
    # s_state may be a copy
    # s_state.loc[:,'s_state_time_cum'] = s_state.loc[:'tau'].cumsum()    
    # TypeError: cannot do slice indexing on <class 'pandas.core.index.Int64Index'> with these indexers [tau] of <class 'str'>
    # s_state.loc[:,'s_state_time_cum'] = s_state['tau'].cumsum()  # Still get SettingWithCopyWarning
    s_state_time_cum = s_state['tau'].cumsum()  # You can just use the copy as-is and make a column..
    # s_state['s_state_partition'] = s_state['s_state_time_cum']/s_state['system_time_end']


    

In [96]:

    

c_state_sums = s_state_traces.sum()
# c_state_sums.sort_values(['tau']) # Multiple columns OK
sums_view = c_state_sums.sort_values('tau', ascending=False)
#print(sums_view.head(10)) # Single column string also OK
sums_view[['tau']].head()  # pretty table view
print(sums_view[['tau']].head())  # prints a 'tau' header
# print(sums_view['tau'].head())  # no column headers


    

    




                       tau
complex_state             
28604          1213.883809
28392           938.379253
95192           576.028348
20476           422.622767
99244           415.129178

In [ ]:

    




    

In [ ]:

    




    

In [ ]: