gwaslab 3.4.37__py3-none-any.whl → 3.4.38__py3-none-any.whl

gwaslab/viz_aux_quickfix.py

@@ -5,7 +5,7 @@ from gwaslab.bd_common_data import get_chr_to_number
 from gwaslab.bd_common_data import get_number_to_chr
 from math import ceil
 
-def _quick_fix(sumstats, chr_dict=get_chr_to_number(), scaled=False, chrom="CHR", pos="POS", p="P", mlog10p="MLOG10P",verbose=True, log=Log()):
+def _quick_fix(sumstats, chr_dict=get_chr_to_number(), scaled=False, chrom="CHR", pos="POS", p="P", mlog10p="MLOG10P",log=Log(), verbose=True):
     '''
     quick sanity check for input sumstats
     '''
@@ -45,10 +45,11 @@ def _quick_fix_p_value(sumstats, p="P", mlog10p="MLOG10P", scaled=False,verbose=
     return sumstats
 
 
-def _quick_fix_mlog10p(sumstats,p="P", mlog10p="MLOG10P", scaled=False, verbose=True, log=Log()):
+def _quick_fix_mlog10p(insumstats,p="P", mlog10p="MLOG10P", scaled=False, log=Log(), verbose=True):
     '''
     drop variants with bad -log10(P) values
     '''
+    sumstats = insumstats.copy()
     if scaled != True:
         if verbose:log.write(" -Sumstats P values are being converted to -log10(P)...")
         sumstats["scaled_P"] = -np.log10(sumstats[p].astype("float64"))
@@ -63,7 +64,7 @@ def _quick_fix_mlog10p(sumstats,p="P", mlog10p="MLOG10P", scaled=False, verbose=
     return sumstats
 
 
-def _quick_fix_eaf(seires, verbose=True, log=Log()):
+def _quick_fix_eaf(seires,log=Log(), verbose=True):
     '''
     conversion of eaf to maf
     '''
@@ -73,7 +74,7 @@ def _quick_fix_eaf(seires, verbose=True, log=Log()):
     return seires.copy()
 
 
-def _quick_fix_chr(seires, chr_dict, verbose=True, log=Log()):
+def _quick_fix_chr(seires, chr_dict,log=Log(), verbose=True):
     '''
     conversion and check for chr
     '''
@@ -84,7 +85,7 @@ def _quick_fix_chr(seires, chr_dict, verbose=True, log=Log()):
     return seires
 
 
-def _quick_fix_pos(seires, verbose=True, log=Log()):
+def _quick_fix_pos(seires,log=Log(), verbose=True):
     '''
     force conversion for pos
     '''
@@ -92,7 +93,7 @@ def _quick_fix_pos(seires, verbose=True, log=Log()):
     return seires
 
 
-def _get_largenumber(*args, log=Log()):
+def _get_largenumber(*args,log=Log(), verbose=True):
     '''
     get a helper large number, >> max(pos)
     '''
@@ -108,7 +109,7 @@ def _get_largenumber(*args, log=Log()):
     return large_number
 
 
-def _quick_add_tchrpos(sumstats, chr="chr", pos="POS", large_number=10000000000, dropchrpos=False, verbose=True, log=Log()):
+def _quick_add_tchrpos(sumstats, chr="chr", pos="POS", large_number=10000000000, dropchrpos=False,log=Log(), verbose=True):
     sumstats["TCHR+POS"] = sumstats["CHR"]*large_number + sumstats["POS"]
     sumstats["TCHR+POS"] = sumstats["TCHR+POS"].astype('Int64')
     if dropchrpos == True:
@@ -117,7 +118,7 @@ def _quick_add_tchrpos(sumstats, chr="chr", pos="POS", large_number=10000000000,
     return sumstats
 
 
-def _quick_merge_sumstats(sumstats1, sumstats2):
+def _quick_merge_sumstats(sumstats1, sumstats2, log=Log(), verbose=True):
     merged_sumstats = pd.merge(sumstats1, sumstats2, on="TCHR+POS", how="outer", suffixes=('_1', '_2'))
     merged_sumstats["CHR"] = merged_sumstats["CHR_1"]
     merged_sumstats["POS"] = merged_sumstats["POS_1"]
@@ -126,7 +127,7 @@ def _quick_merge_sumstats(sumstats1, sumstats2):
     merged_sumstats = merged_sumstats.drop(labels=["CHR_1", "CHR_2", "POS_1", "POS_2"],axis=1)
     return merged_sumstats
 
-def _quick_assign_i(sumstats, chrom="CHR",pos="POS"):
+def _quick_assign_i(sumstats, chrom="CHR",pos="POS",log=Log(), verbose=True):
     # sort by CHR an POS
     sumstats = sumstats.sort_values([chrom,pos])
     # set new id
@@ -158,7 +159,7 @@ def _quick_assign_i(sumstats, chrom="CHR",pos="POS"):
     sumstats["i"] = np.floor(pd.to_numeric(sumstats["i"], errors='coerce')).astype('Int64')
     return sumstats, chrom_df
 
-def _quick_assign_i_with_rank(sumstats, chrpad, use_rank=False, chrom="CHR",pos="POS",drop_chr_start=False,_posdiccul=None):
+def _quick_assign_i_with_rank(sumstats, chrpad, use_rank=False, chrom="CHR",pos="POS",drop_chr_start=False,_posdiccul=None,log=Log(), verbose=True):
     # align all variants on a single axis (i)
     sumstats = sumstats.sort_values([chrom,pos])
     if use_rank is True: 
@@ -218,7 +219,7 @@ def _quick_assign_i_with_rank(sumstats, chrpad, use_rank=False, chrom="CHR",pos=
     sumstats["i"] = np.floor(pd.to_numeric(sumstats["i"], errors='coerce')).astype('Int64')
     return sumstats, chrom_df
 
-def _quick_assign_marker_relative_size(series, sig_level = 5e-8, suggestive_sig_level=5e-6, lower_level=5e-4):
+def _quick_assign_marker_relative_size(series, sig_level = 5e-8, suggestive_sig_level=5e-6, lower_level=5e-4,log=Log(), verbose=True):
     size_series = series.copy()
     size_series[:] = 1
 
@@ -231,7 +232,7 @@ def _quick_assign_marker_relative_size(series, sig_level = 5e-8, suggestive_sig_
     size_series[is_sig_level] = 4
     return size_series
 
-def _quick_assign_highlight_hue(sumstats,highlight,highlight_windowkb, snpid="SNPID",chrom="CHR",pos="POS",verbose=True, log=Log()):
+def _quick_assign_highlight_hue(sumstats,highlight,highlight_windowkb, snpid="SNPID",chrom="CHR",pos="POS",log=Log(), verbose=True):
     to_highlight = sumstats.loc[sumstats[snpid].isin(highlight),:]
     #assign colors: 0 is hightlight color
     for i,row in to_highlight.iterrows():
@@ -243,7 +244,7 @@ def _quick_assign_highlight_hue(sumstats,highlight,highlight_windowkb, snpid="SN
         sumstats.loc[right_chr&up_pos&low_pos,"HUE"]="0"
     return sumstats
 
-def _quick_assign_highlight_hue_pair(sumstats, highlight1, highlight2, highlight_windowkb, chrom="CHR",pos="POS",verbose=True, log=Log()):
+def _quick_assign_highlight_hue_pair(sumstats, highlight1, highlight2, highlight_windowkb, chrom="CHR",pos="POS",log=Log(), verbose=True):
     #assign colors: 0 is hightlight color
     to_highlight1 = pd.DataFrame()
     to_highlight2 = pd.DataFrame()
@@ -271,7 +272,7 @@ def _quick_assign_highlight_hue_pair(sumstats, highlight1, highlight2, highlight
             sumstats.loc[right_chr&up_pos&low_pos,"HUE2"]="0"
     return sumstats, to_highlight1, to_highlight2
 
-def _quick_extract_snp_in_region(sumstats, region, chrom="CHR",pos="POS",verbose=True, log=Log()):
+def _quick_extract_snp_in_region(sumstats, region, chrom="CHR",pos="POS",log=Log(), verbose=True):
     region_chr = region[0]
     region_start = region[1]
     region_end = region[2]
@@ -281,7 +282,8 @@ def _quick_extract_snp_in_region(sumstats, region, chrom="CHR",pos="POS",verbose
     sumstats = sumstats.loc[is_in_region_snp,:]
     return sumstats
 
-def _cut(series, mode,cutfactor,cut,skip, ylabels, cut_log, verbose,lines_to_plot, log):
+def _cut(series, mode,cutfactor,cut,skip, ylabels, cut_log, verbose, lines_to_plot, log):
+    log.write(" -Converting data above cut line...",verbose=verbose)
     if ylabels is not None:
         ylabels = pd.Series(ylabels)
     maxy = series.max()
@@ -340,7 +342,7 @@ def _cut(series, mode,cutfactor,cut,skip, ylabels, cut_log, verbose,lines_to_plo
                 #sumstats.loc[sumstats["scaled_P"]>cut,"scaled_P"] = (sumstats.loc[sumstats["scaled_P"]>cut,"scaled_P"]-cut)/cutfactor +  cut
                 
                 maxy = (maxticker-cut)/cutfactor + cut
-    if verbose: log.write("Finished data conversion and sanity check.")
+    
     return series, maxy, maxticker, cut, cutfactor,ylabels,lines_to_plot
 
 #def _cut_line(level, mode,cutfactor,cut,skip, ylabels, cut_log, verbose, log):
@@ -362,9 +364,11 @@ def _set_yticklabels(cut,
                      font_family,
                      ytick3,
                      ylabels,
-                     ylabels_converted
+                     ylabels_converted,
+                     log=Log(),
+                     verbose=True
                      ):
-    
+    log.write(" -Processing Y tick lables...",verbose=verbose)
     # if no cut
     if cut == 0: 
             ax1.set_ylim(skip, ceil(maxy*1.2) )
@@ -430,7 +434,8 @@ def _set_yticklabels(cut,
     
     return ax1
 
-def _jagged_y(cut,skip,ax1,mode,mqqratio,jagged_len,jagged_wid):
+def _jagged_y(cut,skip,ax1,mode,mqqratio,jagged_len,jagged_wid, log=Log(), verbose=True):
+    log.write(" -Processing jagged Y axis...",verbose=verbose)
     tycut = cut +0.3 #(cut - skip)/ (ax1.get_ylim()[1] - skip) + 0.002
     dy= jagged_len * (cut - skip) 
     x0 =  0

gwaslab/viz_aux_reposition_text.py

@@ -2,7 +2,7 @@ import pandas as pd
 import numpy as np
 from gwaslab.g_Log import Log
 
-def adjust_text_position(positions, yspan, repel_force=0.01, max_iter=100,amode="int",log=Log(),verbose=True):
+def adjust_text_position(positions, yspan, repel_force=0.01, max_iter=100,amode="int",log=Log(),verbose=True, min_factor=None):
     # check the number of variants to annotate 
     #if repel_force>0:
     #    if 1/(repel_force*2 +0.01) < len(positions):
@@ -15,10 +15,12 @@ def adjust_text_position(positions, yspan, repel_force=0.01, max_iter=100,amode=
     if amode=="int":
         step = int(yspan*repel_force) 
     elif amode=="log":
-        min_factor = np.min(positions)
+        if min_factor is None:
+            min_factor = np.min(positions)
         #(1, max) -> (0, log(max)))
-        positions = np.log(positions/min_factor)
+        positions = np.log2(positions/min_factor)
         step = max(positions)*repel_force 
+
     else:
         step = yspan*repel_force 
 
@@ -33,7 +35,8 @@ def adjust_text_position(positions, yspan, repel_force=0.01, max_iter=100,amode=
             if amode=="int":
                 return  np.floor(pd.to_numeric(positions, errors='coerce')).astype('Int64').copy()
             elif amode=="log":
-                return  np.exp(pd.to_numeric(positions, errors='coerce')) * min_factor
+                
+                return  np.power(2, pd.to_numeric(positions, errors='coerce'))* min_factor
             else:
                 return  pd.to_numeric(positions, errors='coerce')
         else:

gwaslab/viz_aux_save_figure.py

@@ -4,23 +4,24 @@ import time
 import os.path
 
 def save_figure(fig, save, keyword, save_args=None, log = Log(), verbose=True):
+    log.write("Start to save figure..." ,verbose=verbose)
     if save_args is None:
         save_args = {}
     if save:
-        if verbose: log.write("Saving plot:")
         if save==True:
             default_path = get_default_path(keyword)
             fig.savefig(default_path, bbox_inches="tight",**save_args)
-            log.write(" -Saved to "+ default_path + " successfully!" )
+            log.write(" -Saved to "+ default_path + " successfully!" ,verbose=verbose)
         else:
             if os.path.exists(save):
                 fig.savefig(save,bbox_inches="tight",**save_args)
-                log.write(" -Saved to "+ save + " successfully! (overwrite)" )
+                log.write(" -Saved to "+ save + " successfully! (overwrite)" ,verbose=verbose)
             else:
                 fig.savefig(save,bbox_inches="tight",**save_args)
-                log.write(" -Saved to "+ save + " successfully!" )
+                log.write(" -Saved to "+ save + " successfully!" ,verbose=verbose)
     else:
-        log.write(" -Skip saving figures!" )
+        log.write(" -Skip saving figure!" ,verbose=verbose)
+    log.write("Finished saving figure..." ,verbose=verbose)
 
 def get_default_path(keyword,fmt="png"):
     path_dictionary = { 

gwaslab/viz_plot_compare_af.py

@@ -72,11 +72,11 @@ def plotdaf(sumstats,
     
 
     sumstats = sumstats.loc[(~sumstats[eaf].isna())&(~sumstats[daf].isna()),[snpid,eaf,daf]+alleles].copy()
-    sumstats.loc[:,daf] = sumstats.loc[:,daf].astype("float")
-    sumstats.loc[:,eaf] = sumstats.loc[:,eaf].astype("float")
+    sumstats[daf] = sumstats[daf].astype("float")
+    sumstats[eaf] = sumstats[eaf].astype("float")
     if verbose: log.write(" -Plotting valriants:" + str(len(sumstats)))
     
-    sumstats.loc[:,"RAF"]=sumstats[eaf] - sumstats[daf]
+    sumstats["RAF"]=sumstats[eaf] - sumstats[daf]
     sns.set_style("ticks")
     fig, (ax1, ax2) = plt.subplots(1, 2,**plt_args)
     ax1.scatter(sumstats["RAF"],sumstats[eaf],label="Non-outlier", **scatter_args)
@@ -119,9 +119,9 @@ def plotdaf(sumstats,
     ax1.set_ylim([0,1])
     
 
-    sumstats.loc[:,"ID"] = sumstats.index
+    sumstats["ID"] = sumstats.index
     
-    to_plot = pd.melt(sumstats,id_vars=['ID'], value_vars=['EAF',"RAF"], var_name='Types', value_name='Allele Frequency')
+    to_plot = pd.melt(sumstats,id_vars=['ID'], value_vars=['EAF',"RAF"], var_name='Types', value_name='Allele Frequency').dropna()
     
     sns.histplot(data=to_plot, x="Allele Frequency", hue="Types", fill=True, ax=ax2, legend = legend2 ,**histplot_args)
     ax2.set_xlabel("Allele Frequency",**font_args)

gwaslab/viz_plot_miamiplot2.py

@@ -38,6 +38,7 @@ from gwaslab.g_Sumstats import Sumstats
 from gwaslab.viz_aux_save_figure import save_figure
 from gwaslab.viz_plot_mqqplot import mqqplot
 from gwaslab.g_version import _get_version
+
 def plot_miami2( 
           path1=None,
           path2=None,
@@ -86,7 +87,7 @@ def plot_miami2(
           log=Log(),
           **mqq_args
           ):
-    
+    log.write("Start to create miami plot {}:".format(_get_version()), verbose=verbose)
     ## figuring arguments ###########################################################################################################
     # figure columns to use
     if scaled == True:
@@ -98,6 +99,7 @@ def plot_miami2(
             cols = ["CHR","POS","MLOG10P"]
         else:
             cols = ["CHR","POS","P"]
+
     if cols1 is None:
         cols1 = cols.copy()
     if cols2 is None:
@@ -151,9 +153,8 @@ def plot_miami2(
     fig_args, scatter_args = _figure_args_for_vector_plot(save, fig_args, scatter_args)
 
     # add suffix if ids are the same
-    id1, id2, mqq_args1, mqq_args2 = _solve_id_contradictory(id0, id1, id2, mqq_args1, mqq_args2)
-    
-    if verbose: log.write("Start to plot miami plot {}:".format(_get_version()))
+    id1_1, id2_2, mqq_args1, mqq_args2 = _solve_id_contradictory(id0, id1, id2, mqq_args1, mqq_args2)
+
     if dpi!=100:
         fig_args["dpi"] = dpi
     if xtickpad is None:
@@ -176,6 +177,7 @@ def plot_miami2(
             titles_pad_adjusted[0]= 1 + titles_pad[0]
         if "anno2" in mqq_args.keys():
             titles_pad_adjusted[1]=  - titles_pad[1]
+
     if merged_sumstats is None:
     ## load sumstats1 ###########################################################################################################
         sumstats1 = _figure_type_load_sumstats(name="Sumstats1", 
@@ -198,8 +200,8 @@ def plot_miami2(
     else:
         cols1[2] += suffixes[0]
         cols2[2] += suffixes[1]
-        sumstats1 = merged_sumstats.loc[:,cols1].copy()
-        sumstats2 = merged_sumstats.loc[:,cols2].copy()
+        sumstats1 = merged_sumstats[cols1].copy()
+        sumstats2 = merged_sumstats[cols2].copy()
 
     ## rename and quick fix ###########################################################################################################
     renaming_dict1 = {cols1[0]:"CHR",cols1[1]:"POS",cols1[2]:"P"}
@@ -217,7 +219,7 @@ def plot_miami2(
     ## create merge index ###########################################################################################################
     sumstats1 = _quick_add_tchrpos(sumstats1,large_number=large_number, dropchrpos=False, verbose=verbose, log=log)
     sumstats2 = _quick_add_tchrpos(sumstats2,large_number=large_number, dropchrpos=False, verbose=verbose, log=log)
-    if verbose: log.write(" -Merging sumstats using chr and pos...")
+    log.write(" -Merging sumstats using chr and pos...",verbose=verbose)
     
     ###### merge #####################################################################################################
     merged_sumstats = _quick_merge_sumstats(sumstats1=sumstats1,sumstats2=sumstats2)
@@ -231,7 +233,9 @@ def plot_miami2(
                                                           drop_chr_start=False)
     
     # P_1  scaled_P_1  P_2  scaled_P_2  TCHR+POS CHR POS    
-    
+    log.write(" -Columns in merged sumstats: {}".format(",".join(merged_sumstats.columns)), verbose=verbose)
+
+
     del(sumstats1)
     del(sumstats2)
     garbage_collect.collect()
@@ -243,13 +247,14 @@ def plot_miami2(
         plt.subplots_adjust(hspace=region_hspace)    
     else:
         fig, ax1, ax5 = figax
-
+    
+    log.write("Start to create Manhattan plot for sumstats1...", verbose=verbose)
     fig,log = mqqplot(merged_sumstats,
                       chrom="CHR",
                       pos="POS",
                       p="P_1",
                       mlog10p="scaled_P_1",
-                      snpid=id1,
+                      snpid=id1_1,
                       scaled=scaled1,
                       log=log, 
                       mode=mode,
@@ -260,15 +265,16 @@ def plot_miami2(
                       _if_quick_qc=False,
                       **mqq_args1
                      )
+    log.write("Finished creating Manhattan plot for sumstats1".format(_get_version()), verbose=verbose)
 
-
+    log.write("Start to create Manhattan plot for sumstats2...", verbose=verbose)
     fig,log = mqqplot(merged_sumstats,
                       chrom="CHR",
                       pos="POS",
                       p="P_2",
                       mlog10p="scaled_P_2",
                       scaled=scaled2,
-                      snpid=id2,
+                      snpid=id2_2,
                       log=log, 
                       mode=mode,
                       figax=(fig,ax5),
@@ -277,7 +283,8 @@ def plot_miami2(
                        _invert=True,
                       _if_quick_qc=False,
                      **mqq_args2)
-    
+    log.write("Finished creating Manhattan plot for sumstats2".format(_get_version()), verbose=verbose)
+
     if same_ylim==True:
         ylim1_converted = ax1.get_ylim()
         ylim2_converted = ax5.get_ylim()
@@ -285,8 +292,6 @@ def plot_miami2(
             ax5.set_ylim(ylim1_converted)
         else:
             ax1.set_ylim(ylim2_converted)
-    
-    
     #####################################################################################################################    
     
     ax5.set_xlabel("")
@@ -337,24 +342,27 @@ def _sort_args_to_12(mqq_args):
     return mqq_args1, mqq_args2
 
 def _solve_id_contradictory(id0, id1, id2, mqq_args1, mqq_args2):
-    if id1 is not None and id2 is not None:
+    if (id1 is not None) and (id2 is not None):
         if id1 == id2:
             id1_1 = id1 + "_1"
             id2_2 = id2 + "_2"
             if "anno" in mqq_args1.keys():
                 if mqq_args1["anno"] == id1:
                     mqq_args1["anno"] = id1_1
-            if "anno" in mqq_args1.keys():
-                if mqq_args1["anno"] == id2:
-                    mqq_args1["anno"] = id2_2
+            if "anno" in mqq_args2.keys():
+                if mqq_args2["anno"] == id2:
+                    mqq_args2["anno"] = id2_2
         else:
             id1_1 = id1
             id2_2 = id2
+    
     if id1 is None:
         id1_1 = id0
+
     if id2 is None:
         id2_2 = id0
-    return id1_1, id2_2, mqq_args1, mqq_args2
+
+    return (id1_1, id2_2, mqq_args1, mqq_args2)
 
 def _figure_args_for_vector_plot(save, fig_args, scatter_kwargs ):
     if save is not None: