cleaning old code

plots/plotreing.py

@@ -11,53 +11,7 @@ dict_ops = {'size':['Size','size'],'successrate':['Succes rate','srate'],'f-meas
 
 sep = os.environ.get("sep", " ") # default to space
 
-#### All this is legacy code for the first Notebook generation
-def plot_average_curve_for_list_of_patterns( data, label, col, ops, maxval, nbgames, pattern_list ):
-    fig = plt.figure(figsize=(12,3))
-    ax = plt.gca()
-    linestylelist = ['solid', 'dashed', 'dotted', 'dashdot']
-    ind = 0
-    for pattern, ls in zip( pattern_list, linestylelist ):
-        plt.gca().set_prop_cycle(None)
-        for op in ops:
-            data[ind].iloc[1::100].plot(kind='line',y=col, label=op, ax=ax, linestyle=ls, linewidth=1.0) #, marker='*', color='blue'
-            ind += 1
-    ax.set_xlabel('games')  # Add an x-label to the axes.
-    ax.set_xlim([0,nbgames])
-    ax.set_ylabel(label)  # Add a y-label to the axes.
-    ax.set_ylim([0,maxval])
-    #ax.set_ylim([0,100])
-    ax.spines['top'].set_visible(False) # suppress top and right bars
-    ax.spines['right'].set_visible(False)
-    #
-    #ax.set_title(label+" for "+pattern)  # Add a title to the axes.
-    ax.set_title(label)  # Add a title to the axes.
-    #ax.legend(loc='center', bbox_to_anchor=(0.5, -0.05), shadow=True, ncol=2)  # Add a legend.
-    ax.legend() # ncol=2
-    #
-    plt.show()
-    
-# Returns an ordered array of the data for a list of patterns
-def prepare_pattern_list_simple_data ( ops, nbgames, pattern_list, results_directory, observed_variables ):
-    dataArray = []
-    for pattern in pattern_list:
-        for op in ops:
-            dataArray.append( load_norm_data( pattern.replace( "%%", op ), results_directory, observed_variables ) )
-    return dataArray
-
-def print_data_series( pattern_list, nbiter, results_directory, observed_variables ):
-    allops = ['delete','replace','refine','add','addjoin','refadd']
-    dataframes = prepare_pattern_list_simple_data( allops, nbiter, pattern_list, results_directory, observed_variables )
-    # here it would be good to scrap the MAXSIZE. This can be done in the data
-    plot_average_curve( dataframes, "Size", 'size', allops, 120, nbiter, pattern_list )
-    plot_average_curve( dataframes, "Success rate", 'srate', allops, 1, nbiter, pattern_list )
-    plot_average_curve( dataframes, "Precision", 'prec', allops, 1, nbiter, pattern_list )
-    plot_average_curve( dataframes, "F-measure", 'fmeas', allops, 1, nbiter, pattern_list )
-    plot_average_curve( dataframes, "Recall", 'rec', allops, 1, nbiter, pattern_list )
-
-
-
-#### All this is new code for Notebook generation
+#### New code for Notebook generation
 
 # Returns the prefix of the files in the results directory of the experience 'label'
 # prefix could be label+nbAgents+nbIterations or just nbAgents+nbIterations
@@ -128,42 +82,6 @@ def prepare_data_for_labels(labelpatts,base_directory,observed_variables):
         dataDict[filepattern] = df
     return(dataDict)
 
-# Prepare data for all plots
-#Returns a dictionary where the keys are combinations of an experience+pattern
-def prepare_all_dataOld(allplots,results_directory,labelexp,prefix,ops,observed_variables):
-    dataDict = {}
-    for group in allplots:
-        #a group: [[['Size', 'Games', '10000', '120'], ['20180601-NOOR', 'clever-nr-gen'], ['20180529-NOOR', 'clever-nr']], [['Size', 'Games (im80)', '10000', '120'], ['20180601-NOOR', 'clever-nr-im80-gen'], ['20180530-NOOR', 'clever-nr-im80']]]
-        for plot in group:
-            #a plot: [['Size', 'Games', '10000', '120'], ['20180601-NOOR', 'clever-nr-gen'], ['20180529-NOOR', 'clever-nr']]
-            for filepair in plot[1:]:
-                #a filepair: ['20180601-NOOR', 'clever-nr-gen']
-                #filepair[0] can be the current experience or other experience
-                #filepair[1] is a pattern
-                filepattern = filepair[0]+filepair[1]
-                if not (filepattern in dataDict):
-                    #This is for manage cases where there is no pattern, files concern just operators
-                    opPattern = '%%-'
-                    if filepair[1] == '':
-                        opPattern = '%%'                        
-                    #We look for the directory containing result files and also for the prefix of the files in that directory
-                    # LABEL+NBAGENTS+ITERATIONS or just NBAGENTS+ITERATIONS
-                    filesprefix = prefix
-                    resdir = results_directory
-                    if not filepair[0] == labelexp:
-                    #if (not filepair[0] == labelexp) and (labelexp != '20170529-NOOR'):
-                        local_directory = os.getcwd()
-                        resdir = local_directory+'/../'+filepair[0]+'/results/'
-                        filesprefix = getPrefix(resdir,filepair[0],'tsv')
-                        #This also works for resdir
-                        #resdir = results_directory+'/../../'+filepair[0]+'/results/'
-                    #display("Will prepare dataframe for "+filesprefix+filepair[1]+" in "+resdir)
-                    if (labelexp == '20170208-NOOR'):
-                        ops = ['add','del','repl']
-                    df = prepare_pattern_data(ops,filesprefix+opPattern+filepair[1],resdir,observed_variables)
-                    dataDict[filepattern] = df
-    return(dataDict)
-    
 # Prepare data needed for all plots
 # Returns a dictionary where the keys are combinations of an experience+pattern
 def prepare_all_data(expPatterns,results_directory,labelexp,prefix,ops,observed_variables):
@@ -237,7 +155,7 @@ def plot_datafiles_curve( alldata, indexes, ops, ylabel, xlabel, xmax, ymax, col
     #
     plt.show()
     
-def plot_average_curve( alldata, indexes, ylabel, xlabel, xmax, ymax, col, ops, nbgames ):
+def plot_average_curve( alldata, indexes, ylabel, xlabel, xmax, ymax, col, ops ):
     fig = plt.figure(figsize=(12,3))
     ax = plt.gca()
     linestylelist = ['solid', 'dashed', 'dotted', 'dashdot']
@@ -279,7 +197,7 @@ def print_datafiles( alldf, plotinfo, labelpatts, cols, nbiter ):
     #plot_datafiles_curve_data( alldf, indexes, cols )
     plot_datafiles_curve( alldf, indexes, cols, plotlabel, xlabel, xmax, ymax, col, nbiter )
 
-def print_one_series( alldf, plotinfo, patterns, operators, nbiter ):
+def print_one_series( alldf, plotinfo, patterns, operators ):
             # plotinfo: ['Size', 'Games', '10000', '120']
             # patterns: [['20180601-NOOR', 'clever-nr-gen'], ['20180529-NOOR', 'clever-nr']]
             #print("Will build plot for ")
@@ -294,14 +212,9 @@ def print_one_series( alldf, plotinfo, patterns, operators, nbiter ):
             xlabel = plotinfo[1]
             xmax = int(plotinfo[2])
             ymax = int(plotinfo[3])
-            plot_average_curve( alldf, indexes, plotlabel, xlabel, xmax, ymax, col, operators, nbiter )
+            plot_average_curve( alldf, indexes, plotlabel, xlabel, xmax, ymax, col, operators )
             
 ### Print all plots of an experience extracted from the latexfile
-#def print_all_series( label, latexfile, prefix, nbiter, results_directory, operators, observed_variables ):
-    #allplots = getPlots(latexfile,label,'[0-9]-[0-9]+-')
-    #print(plots)
-#def print_all_series( label, allplots, prefix, nbiter, results_directory, operators, observed_variables ):
-#    alldf = prepare_all_data(allplots,results_directory,label,prefix,operators,observed_variables)
 def print_all_series( allplots, alldf, operators, nbiter ):
     for group in allplots:
         for plot in group: