def getTaxonAbundance(self, level, taxon):
self.r.getTaxonMap(level)
df_sub_data_nmds= self.r.getTaxonAbundance(self.df_data_nmds, level, taxon)
df_sub_data_diversity= self.r.getTaxonAbundance(self.df_data_diversity, level, taxon)
sub_data_nmds= pandas_df_to_r_df(self.df_sub_data_nmds.T)
sub_data_diversity= pandas_df_to_r_df(self.df_sub_data_diversity.T)
return sub_data_nmds, sub_data_diversity
def transformDFByRFunction(self, df, function, library="", *args):
r_df= pandas_df_to_r_df(df)
if library != "":
import_=r.r("library(%s)"%library)
function_=r.r(function)
r_updated_df= function_(r_df, *args)
return r_matrix_to_dataframe(r_updated_df)
def plotCommunityResponse(self, diversity_data, metadata):
title= self.r.initRVariable('title', self.analysis_type)
metadata= metadata.ix[diversity_data.index]
r_diversity_data= pandas_df_to_r_df(diversity_data)
r_dis_matrix= self.r.fPlotCommunityResponse(r_diversity_data)
dis_matrix= r_matrix_to_dataframe(r_dis_matrix)
### build a new dataframe with triclosan values
A={int(b):a for a,b in zip(diversity_data.T ,set(list(dis_matrix['Var1'])))}
new_df= DataFrame(columns= ["Var1","Var2","value","Triclosan1","Triclosan2"], index= dis_matrix.index)
for index in dis_matrix.index:
try:
new_df['Var1'][index]= A[dis_matrix['Var1'][index]]
new_df['Var2'][index]= A[dis_matrix['Var2'][index]]
new_df['Triclosan1'][index]= self.r.df_metadata.ix[new_df['Var1'][index]]['Triclosan']
new_df['Triclosan2'][index]= self.r.df_metadata.ix[new_df['Var2'][index]]['Triclosan']
new_df['value'].loc[index]= dis_matrix['value'].loc[index]
except Exception,e:
print e
continue
def plotIndex(self, diversity_data, metadata, metadata_type, r=False):
#df= r_matrix_to_dataframe(diversity_data)
#df_merged_data= df.join(self.df_metadata)
#pdb.set_trace()
metadata= metadata.ix[diversity_data.index]
r_diversity_data= pandas_df_to_r_df(diversity_data)
indice_data= self.r.fBuildDiversityIndices(r_diversity_data)
#pdb.set_trace()
#index_data= map(np.array, indice_data)
Indice_Data= dict(zip(("chao1", "shannon", "richness", "pielou"), indice_data))
for indice, Index in self.r.divIndexes.iteritems():
indice_data= Indice_Data[indice]
ind= self.r.divIndexes[indice]
ind.data= indice_data
try:
ind_data= list(ind.data)
names= diversity_data.index
s1= Series(index= names, data= ind_data )
s2= metadata
df_indice= concat([s2,s1], axis=1)
df_indice.columns= ['metadata', indice]
except:
pass
if r:
self.r.fPlotDiversityIndex(ind.type_short, ind.type_long, metadata, ind.data, ind.title)
else:
self.plotDiversityIndex(ind.type_short, ind.type_long,metadata_type, df_indice, ind.data)
if indice == "richness":
fRichness= os.path.join(self.analysis_dir, 'richness.tsv')
ind_data= list(ind.data)
names= list(ind.data.names)
s1= Series(index= names, data= ind_data )
s2= metadata
df_indice= concat([s2,s1], axis=1)
df_indice.columns= ['metadata', indice]
df_indice.sort('metadata').to_csv(fRichness, sep='\t', index_label="Sample")
def applyRFunctionToDF(self, df, function, library="", *args):
r_df= pandas_df_to_r_df(df)
if library != "":
import_=r.r("library(%s)"%library)
function_=r.r(function)
return function_(r_df, *args)
def plotNMDS(self, nmds_data, metadata, analysis_type):
r_nmds_data= pandas_df_to_r_df(nmds_data)
r_analysis_type= self.r.initRVariable('analysis_type', analysis_type)
r_nmds= self.r.fMetaMDS(r_nmds_data, r_analysis_type)
r_scores= self.r.fMDSScores(r_nmds)
stress= list(r_nmds.rx('stress')[0])[0]
scores= r_matrix_to_dataframe(r_scores)
plt.figure()
dim1= scores["NMDS1"]
dim2= scores["NMDS2"]
metadata= metadata.ix[nmds_data.index]
c_data = plt.cm.jet(np.log10(metadata))
cm = plt.cm.get_cmap('Greys')
# adding a negligible value to the controls so that logs can be meaningful
metadata[metadata == 0.0] = metadata[metadata == 0.0] +0.1
sc = plt.scatter(dim1,dim2, c= np.log10(metadata), vmin=min(np.log10(metadata)), vmax=max(np.log10(metadata)),s=100, cmap=cm)
cb= plt.colorbar(sc, shrink= 0.6, aspect= 10, pad= 0.1, norm=matplotlib.colors.LogNorm())
cb.ax.tick_params(labelsize='x-small',top=False)
cb.set_label('Triclosan concentration [nM]', labelpad=-70)
labels= np.sort(np.array(list(set(metadata))))
ind = np.log10(labels)
ind_labels= map(str, labels)
ind_labels[0] = "0.0"
cb.set_ticks(ind)
cb.set_ticklabels(ind_labels)
ax= plt.gca()
ax.set_xlabel('NMDS1')
ax.set_ylabel('NMDS2')
ax.set_title('NMDS Ordination of %s OTU Abundance' %self.analysis_type)
left, width = 1.05, .5
bottom, height = 0.05, .5
ax.grid(True, color="gray", alpha= 0.5)
ax.text(left, bottom, 'Stress: %.2g'% stress, fontsize= 11,
horizontalalignment='left', verticalalignment='bottom',
transform= ax.transAxes)# bbox={ 'facecolor':'white','pad':10})
plt.savefig( os.path.join(self.analysis_dir,'nmds.png'))
annotations= map(str, list(metadata))
for label, x, y in zip(annotations, dim1, dim2):
plt.annotate(
label,
xy = (x, y), xytext = (random.randint(-50,-1), random.randint(1,50)),
textcoords = 'offset points', ha = 'right', va = 'bottom',
bbox = dict(boxstyle = 'round,pad=0.8', fc = 'yellow', alpha = 0.2),
arrowprops = dict(arrowstyle = '->', connectionstyle = 'arc3,rad=0'),
fontsize= 10)
#"""
plt.savefig( os.path.join(self.analysis_dir, 'nmds_with_sample_names.png'))
plt.close()
def plotRarefaction(self, nmds_data):
r_analysis_type= self.r.initRVariable('analysis_type', self.analysis_type)
r_analysis_dir= self.r.initRVariable('analysis_dir', self.analysis_dir)
r_nmds_data= pandas_df_to_r_df(nmds_data)
rarefaction= self.r.fPlotRarefaction(r_nmds_data, r_analysis_type, r_analysis_dir)
