I just install pycomplexheatmap==1.6.5
pyComplextHeatmap used (col_split=2, row_split=3) as an clustering example in its document, yet when I add col_split to 3 or 3+, it will make mistake easily

cm = ClusterMapPlotter(data=df_heatmap, top_annotation=col_ha, col_split=3, row_split=6, col_split_gap=0.5,
row_split_gap=1,label='values',row_dendrogram=True, col_dendrogram=True,show_rownames=False,show_colnames=False,
tree_kws={'row_cmap': 'Dark2', 'col_cmap': 'Dark2'},cmap='Spectral_r', legend_gap=5,legend_hpad=2,legend_vpad=5)

ValueError: The number of observations cannot be determined on an empty distance matrix.

So this situation appears in my data too, and what's even worse, some branches are in the wrong place.

cm = ClusterMapPlotter(data=data, label='values', 
                       col_split=2, row_split=3,
                       # row_cluster_method="average", row_cluster_metric="correlation",
                       # col_cluster_method="average", col_cluster_metric="correlation",
                       # tree_kws={'row_cmap': 'Dark2'},
                       row_dendrogram=True, col_dendrogram=True, cmap='Spectral_r', show_rownames=True, show_colnames=True
                       )

Hi! Is it possible to make the scatterplot with custom colors? I tried to give my color list as colors, but that argument accepts only single color, tried providing my color list as ListedColormap(list_of_colors), but cmap accepts only strings as an argument. Is there any way?

Hi

Is there a way to center the color bar around a specific number? For instance, centering at 0, without having to Z_score or scale by rows/columns?

Thanks

Hi! Could you please advice how to add ylim to anno_scatterplot?

It seems the functions of row_split and col_split are a bit weird. See the example below:

import pandas as pd
from PyComplexHeatmap import *
data = pd.DataFrame([[0,1],
                     [1,0],
                     [0,1],
                     [1,0],
                     [0,1]])
ClusterMapPlotter(data=data, linewidth=1, row_dendrogram=True, row_cluster=True)

After adding row_split=2, one should expect to see the first two rows in the above image separated from the rest three rows. However, it looks like this is not the case and ClusterMapPlotter returns something different.

import pandas as pd
from PyComplexHeatmap import *
data = pd.DataFrame([[0,1],
                     [1,0],
                     [0,1],
                     [1,0],
                     [0,1]])
ClusterMapPlotter(data=data, linewidth=1, row_dendrogram=True, row_cluster=True, row_split=2, row_split_gap=5)

Do you have any opinion on this?

I would like to standardize the code structure to adapt the code to common style guidelines.

• #47, more information on black
• set up pre-commit hooks
• adapt code according to pylint, pep8

I understand that this is a larger change, however I believe that it will improve the readability of the codebase and makes it easier maintainable.

heatmap 的柱状图的范围不正确，应该是0～0.5， 而且理论上图例不应该显示0～0.5 颜色变化

I'm a pathology image analysis researcher and use your library for creating heatmaps.

I would appreciate if you could add a feature allowing images to be displayed as labels for each feature (row) in the heatmap, instead of text labels.
This enhancement would greatly aid in visual data interpretation in pathology image analysis.

Thank you for considering my suggestion.

I want to plot 22 groups on top of my heatmap with light and dark grey.
For the colors argument I'm using the following dictionary:

{1: '#999993',
 2: '#eeeee4',
 3: '#999993',
 4: '#eeeee4',
 5: '#999993',
 6: '#eeeee4',
 7: '#999993',
 8: '#eeeee4',
 9: '#999993',
 10: '#eeeee4',
 11: '#999993',
 12: '#eeeee4',
 13: '#999993',
 14: '#eeeee4',
 15: '#999993',
 16: '#eeeee4',
 17: '#999993',
 18: '#eeeee4',
 19: '#999993',
 20: '#eeeee4',
 21: '#999993',
 22: '#eeeee4'}

And then plot the heatmap like this:

plt.figure(figsize=(10, 5))


cols = HeatmapAnnotation(chromosome = anno_simple(cut_chroms.chromosome,add_text=True, colors = colors, text_kws ={"color": "black"} )
                         ,plot_legend = True,axis = 1,plot = False)
cm1 = ClusterMapPlotter(data=hm,row_dendrogram=True,show_rownames=False,show_colnames=False,top_annotation = cols,
                       row_cluster_metric = "euclidean",col_cluster = False,
                      row_cluster = True, label='Expression')

plt.show()

As you see, in the initial dictionary group 16 was supposed to be light grey (unlike 15 and 17) and you can also see that in the legend it's light grey, but in the heatmap it's dark grey. Could you please suggest what might be an issue?

Try to save results from ClusterMapPlotter as pdf, but the heatmap is a figure not in vector.

`plt.figure(figsize=(6, 30))
cm = ClusterMapPlotter(data=data.loc[:,df_col.index.tolist()],
top_annotation=col_ha,show_rownames=True,
row_cluster=True,col_cluster=False,
label='on/off',legend_gap=7,
cmap='Purples',rasterized=True)

plt.savefig(root + "test.pdf")
plt.show()
`

Hi Wubin,

I noticed that PyComplexHeatmap currently treats categorical data the same way as continuous data, presenting the legend as a color bar instead of categorical labels. Do you have any plans to improve this feature? I came across a great repository called catheat (https://github.com/schlegelp/catheat) that uses seaborn to handle categorical data. It would be fantastic to see this integrated into PyComplexHeatmap.

Is your feature request related to a problem? Please describe.
Desire to use col_split and col_split_order to specify my column splits and the order in combination with clustering between groups.

Describe the solution you'd like
I would like to control the col_split_order after cluster_between_groups has been used to do the mean clustering of my groups specified with col_split.

Describe alternatives you've considered
None, but I could use cluster_between_groups and use a vector graphics program for instance to reorder the groups which would be a bit messy.

Additional context

MatplotlibDeprecationWarning: The get_cmap function was deprecated in Matplotlib 3.7 and will be removed two minor releases later. Use matplotlib.colormaps[name] or matplotlib.colormaps.get_cmap(obj) instead

https://matplotlib.org/stable/users/prev_whats_new/whats_new_3.7.0.html

Hello! After using the row_split parameter to group the row items, I want to use the row_split_order parameter to set the group order, but it does not work. What I should input for the row_split_order parameter? Bool, group names list or row item index? Thanks!

The picture does not display the full column name by default. I guess it may be due to occlusion (should all column names be displayed even if they are directly occluded?). I now solve this problem by increasing the width of the picture.

When trying to plot two clustermaps that share one axis, PyComplexHeatmap.clustermap.composite() enforces both column and indexes to be the same as the main cluster map.

This is enforced in the following code:

for i, cm in enumerate(cmlist):
        if i == main:
            continue
        gs1 = gs[i, 0] if axis == 0 else gs[0, i]
        cm.plot(ax=axes[i], subplot_spec=gs1, row_order=cm_1.row_order, col_order=cm_1.col_order)
        for L in cm.legend_list:
            if L[1] not in legend_names:
                legend_names.append(L[1])
                legend_list.append(L)
        w = ax.figure.get_window_extent().width * cm.label_max_width / cm.ax.figure.get_window_extent().width
        if w > label_max_width:
            label_max_width = w

The parameters row_order and col_order in cm.plot should be given dynamically depending on whether axis = 0 or axis = 1

The use of the proprietary Microsoft font Arial generates a warning for each time a label is printed out on any system that doesn't have that font installed. Choice of specific fonts (especially proprietary ones) is best left to the user.

I would recommend only specifying the "Sans-serif" font family and then letting the user's matplotlib installation use whichever applicable font of that type they have installed instead of specifying a specific font.

Is there a way to adjust the width and height of the heatmap?
I have tried to create a figure with plt.figure(figsize=(3,5)) before calling ClusterMapPlotter to make the heatmap skinnier. However, this forces the legend to get cut off, even when I save the figure with bbox_inches='tight'.

Hi,

The images you use in your wiki page seem to be GitHub theme dependent. I had to change my GitHub theme to view them properly. My default them is "Dark default". When I view the wiki page with that theme, this is what I see:

But when I switch to the "Light default" theme, I see everything properly:

Dear Wubin,

Thank you very much for creating this incredibly useful package. I have three questions regarding some settings.

1. How can I set the frameon to False for categorical legends?
2. How can I set the colorbar attributes such as aspect or orientation?
3. Apparently if row_split or col_split is not set to None axes are visible, how one in this case can turn them off?

Thank you in advance for your time.

version 1.6.1
matplotlib ver. 3.8.0
numpy ver. 1.26.1
pandas ver. 2.1.1
I ran the example code and the resulting figure only showed part of the annotation.

I'm wondering how to remove the arrows from anno_label and just show the labels alone? I've tried changing values for the arrowprops argument and relpos but whatever I do I can't get rid of the lines. I can't find anything in the docs or the source to help but would love to be able to do this.
Thanks!

df_bar = pd.DataFrame(np.random.uniform(0, 10, (10, 3)), columns=['TMB1', 'TMB2', 'TMB3']) print(df_bar) plt.figure(figsize=(6, 6)) col_ha = HeatmapAnnotation(bar=anno_barplot(df_bar,legend=True), plot=True,legend=True,legend_gap=3, legend_width=20) plt.show()

Please provide a tag for the 1.3.1 release. Tag helps us in easier packaging of this package in the other ecosystems.

The register_cmap function was deprecated in Matplotlib 3.7 and will be removed two minor releases later. Use matplotlib.colormaps.register(name) instead.

How to set a title of heatmap？

Dear wubin:

Thank you so much for creating this wonderful package , I have a question about the legend visualization . In my datasets , the legend was not fully visualized , see the code below , and the datasets were uploading here .

# 横注释
col_ha = HeatmapAnnotation(#label=anno_label(obs.loc[:,'T/NK_type'], colors=type_color_TNK, merge=True,rotation=15), # 设置是否有对T/NK_type的文字注释
               T_NK_type=anno_simple(obs.loc[:,'T/NK_type'], colors=type_color_TNK, add_text=False, legend=True, height=8),axis=1, # 绘制T/NK_type注释
               T_NK_celltype=anno_simple(obs.loc[:,'T/NK_celltype'], colors=type_color_rna, add_text=False, legend=True, height=8), # 绘制T/NK_celltype注释
               plot=False,legend=False,legend_gap=3,legend_hpad=1,legend_width=3
                           )
# 纵注释
row_ha = HeatmapAnnotation(#label=anno_label(obs.loc[:,'T/NK_type'], colors=type_color_TNK, merge=True,rotation=15), # 设置是否有对T/NK_type的文字注释
               Gene=anno_simple(var.loc[:,'celltype'], colors=type_color_rna, add_text=False, legend=True, height=8),axis=0, # 绘制T/NK_type注释
               plot=False,legend=False,legend_gap=3,legend_hpad=1,legend_width=3
                           )
plt.figure(figsize=(8, 16))
# 绘制注释
# 绘制基因表达图
cm = ClusterMapPlotter(data=Gene_Expression.T, #z_score = True,
                       top_annotation = col_ha, # 设置上注释
                       left_annotation = row_ha, # 设置左注释
                       row_cluster = False, # 设置禁止自动聚类！！！
                       col_cluster = False, # 设置禁止自动聚类！！！ 这个逼函数聚类参数默认是True
                       row_names_side = 'right',  # 基因名放右边
                       col_split = obs.loc[:,'T/NK_type'] , # 设置纵轴分割
                       row_split = var.loc[:,'celltype'] , # 设置横轴分割
                       label='values',  # 热图颜色设置
                       row_dendrogram = False,
                       show_rownames = True,
                       show_colnames = False,
                       cmap ='PuBu' , # Values的色卡
                       legend = False, # 不画图例
                      # legend_hpad = 0.5,#热图和图例之间间隙
                      # legend_vpad = 0.5,
                      # legend_side = 'left', # 图例放左边
                       row_split_gap = 2, # 横向分割间距
                       col_split_gap = 0.4, # 纵向分割间距
                    # tree_kws={'row_cmap': 'Dark2'}
                    )

# [Gene_Expression.csv](https://github.com/DingWB/PyComplexHeatmap/files/10553588/Gene_Expression.csv)
plt.savefig("Figure2_marker.png",dpi=300,bbox_inches = 'tight')
plt.savefig("/content/drive/MyDrive/Gastric/Figure/Figure2/Figure2_marker.png",dpi=300,bbox_inches = 'tight')
plt.show()

datasets:
var.csv
obs.csv
Gene_Expression.csv

Hi，How to set the precision of the number, such as keeping 4 significant digits

Hi, PyComplexHeatmap provides Python users with convenient heatmap painting tools. I want to display variants in a population using oncoPrintPlotter. I want to show SampleID in rows and VariantID in columns. Here is the code:

cols = ['0/0','0/1','1/1','./.']
colors = ['#552365','#2e9a95','#f7eb3c','#050707']

plt.figure(figsize=(18,20))
op = oncoPrintPlotter(data = stack_df,x = 'VariantID',y = 'SampleID',
                    values = cols,colors = colors,label = 'Genotype',
                    legend_hpad = 0,show_rownames = True,show_colnames = True,
                    width = 1.8,row_gap = 0)
plt.savefig(query_region_variant_plot,bbox_inches='tight')
plt.show()

The heatmap gets the wrong variant split shown in the red circles:

I set the VariantID as the x-axis. The ticklabels in the figure cannot match the heatmap columns. And the column number in the heatmap is obviously great than the VariantID count.

The input stack_df contains the genotype information for each variant :

Here I show the input file and the output figure.
input data
variant_stack_record.txt
output figure
variant_plot.pdf

cm =ClusterMapPlotter(data=df_heatmap, z_score=0)

cm.data2d will not return clustered matirx but return a matrix with original row/colunm order. So how to get a clustered matirx like using sns.clustermap.
Thanks.

df_bar = pd.DataFrame(np.random.uniform(0, 10, (10, 3)), columns=['TMB1', 'TMB2', 'TMB3']) print(df_bar) plt.figure(figsize=(6, 6)) col_ha = HeatmapAnnotation(bar=anno_barplot(df_bar,legend=True), plot=True,legend=True,legend_gap=3, legend_width=20) plt.show()

Is there a way to center custom annotation vertically ? In this example in the documentation, the asterisks are not centered vertically.

Dear Developer,

For show_rownames= True, could you please help how to only show those row name we are interested in? Thanks!

In the complexheatmap of R, there is a parameter called cluster_columns or clusters_rows allows for the transfer of reordered Dendrogram data (such as using Dendsort packedge in R), which changes the order of the dendrogram in the heatmap and allows it to be displayed. I think this is a very useful function. In pycomplexheatmap, it seems that the order of rows or columns can be passed in, but the dendrogram cannot be displayed anymore. I think it may be possible to add a parameter cluster_ columns or clusters_ rows to pycomplexheatmap as well, an optimized link matrix can be passed in, which can change the order and corresponding clustering tree in the heatmap.

when set *_dendrogram=True
cm = ph.ClusterMapPlotter(
data=z_df, top_annotation=col_ha, left_annotation=row_annotation,
label='z-score', show_rownames=False, show_colnames=True, col_cluster=True,
row_cluster=True, col_split_gap=3,
col_dendrogram=True, row_dendrogram=True, legend_width=50, verbose=0)

when set *_dendrogram=False

I see examples to show a cluster graph, but what if we need the derived clusters? What object should be called?

Discussed in #29

row_ha = HeatmapAnnotation(
                           Category=anno_simple(df_row.Category,cmap='Set1',
                                           add_text=False,legend=False),
                           label=anno_label(df_row.Category, merge=True,rotation=0),
                           axis=0,verbose=0,label_kws={'rotation':45,'horizontalalignment':'left'})

col_ha = HeatmapAnnotation(
                           Dataset=anno_simple(df_col.Dataset,cmap='Set1',legend=False,add_text=True),
                           Correlation=anno_simple(df_col.Correlation,cmap='Dark2',legend=False,add_text=True),
                           verbose=0,label_side='left',label_kws={'horizontalalignment':'right'})

plt.figure(figsize=(5, 8))
cm = DotClustermapPlotter(data=data, x='ID',y='Term',value='-log10(Pval)',c='-log10(Pval)',s='odds_ratio',
                          hue='EnrichType', row_cluster=False,col_cluster=False,
                          cmap={'Enrich':'RdYlGn_r','Depletion':'coolwarm_r'},
                          colors={'Enrich':'red','Depletion':'blue'},
                          #marker={'Enrich':'^','Depletion':'v'},
                          top_annotation=col_ha,right_annotation=row_ha,
                          col_split=df_col.Dataset,row_split=df_row.Category, col_split_gap=0.5,row_split_gap=1,
                          show_rownames=True,show_colnames=False,row_dendrogram=False,
                          verbose=0,legend_gap=7,alpha=0.8)
#plt.savefig("dotHeatmap1.pdf",bbox_inches='tight')
plt.show()

Describe the bug
Missing custom annot (stars indicating statistical significance) when using clustermap.composite. Individual ClusterMapPlotter calls successfully displayed the heatmap annotation.

Environment
Python 3.11.7
PyComplexHeatmap 1.6.4
matplotlib 3.7.1

I have trouble using HeatmapAnnotation. I keep getting empty annotation... any thought?

    PathwayModule = [
        'WPM' if node in list(set(sub_mat.columns) & set(gmt[geneset_name])) else 'BPM' 
        for node in sub_mat.columns
    ]
    
    fig, ax = plt.subplots(figsize=(2 * rel,1.6 * rel))
    
    col_colors_dict={
        'WPM':'#1E93AE',
        'BPM':'#FF9C00'
    }    
    col_ha = ch.HeatmapAnnotation(
        label=ch.anno_label(
            pd.Series(PathwayModule),merge=True,rotation=90,extend=True,
            adjust_color=True,luminance=0.75,
            colors=col_colors_dict,
            relpos=(0.5,0)
        ),
        PathwayModule=ch.anno_simple(
            pd.Series(PathwayModule),
            colors=col_colors_dict,
        ), axis=1,
        label_side='right',
        verbose=0,
        # plot=False,
        legend=False
    )
    plot_cm = ch.ClusterMapPlotter(
        sub_mat,
        legend_gap=5,legend_width=5,legend_hpad=2,legend_vpad=5,
        right_annotation=col_ha,
        col_dendrogram=True,
        row_dendrogram=True,
        vmin=-1, vmax=1,
        cmap = 'BlueYellow',
        verbose=False,
        legend=False,
        # plot=False,
        # show_rownames=True,show_colnames=True
    )
    

Will the Pycomplexheatmap support Kmeans clustering when using ClusterMapPlotter function. I just mean that first use Kmeans to form different goups, then run hierarchical clustering within each group, finally hierarchically cluster different groups. The R package complexheatmap support the process.
Thanks!

Is your feature request related to a problem? Please describe.
A clear and concise description of what the problem is. Ex. I'm always frustrated when [...]

Describe the solution you'd like
A clear and concise description of what you want to happen.

Describe alternatives you've considered
A clear and concise description of any alternative solutions or features you've considered.

Additional context
Add any other context or screenshots about the feature request here.

I am wordering can I center the colorbar with white representing 0 values ??
plt.figure(figsize=(10, 4))

cm = ClusterMapPlotter(data=df_heatmap.T,
                       col_cluster=False,row_cluster=False,
                       row_split=df.Group,col_split=2,
                       col_split_gap=0.5,row_split_gap=0.8,
                       label='values',row_dendrogram=True,
                       show_rownames=False,show_colnames=True,
                       tree_kws={'row_cmap': 'Set1'},verbose=0,legend_gap=5,
                       cmap='bwr',xticklabels_kws={'labelrotation':90,'labelcolor':'black'},
                       #legend_kws={'vmin':-2,'vmax':2}
                       
                      )
#plt.savefig("example0.pdf", bbox_inches='tight')
plt.show()

When other packages are imported, such as seaborn and set some kind of style, there would be a unexpected grid in heatmap

import os,sys
%matplotlib inline
import matplotlib.pylab as plt
import pickle
plt.rcParams['figure.dpi'] = 100
plt.rcParams['savefig.dpi']=300
plt.rcParams['font.family']='sans serif'
plt.rcParams['font.sans-serif']='Arial'
plt.rcParams['pdf.fonttype']=42
import pandas as pd
from PyComplexHeatmap import *
import PyComplexHeatmap as pch
print(pch.__version__)
import scanpy
#import seaborn as sns
# sns.set_style("whitegrid")
# sns.set_style("white")

df_row=pd.read_csv("df_row.csv",index_col=0)
df_row.spatial_domain=df_row.spatial_domain.astype(str)
df_col=pd.read_csv("df_col.csv",index_col=0)
df_col.chromosome=df_col.chromosome.astype(str)
data=pd.read_csv("cnv_mtr_filtered.csv",index_col=0)
df_col=df_col.sample(500)
data=data.loc[:,df_col.index.tolist()]
data

plt.figure(figsize=(10, 5))


right_annotation = HeatmapAnnotation(
                **,
                       axis=0)

top_annotation = HeatmapAnnotation(
    
                label=anno_label(df_col['chromosome'],merge=True, rotation=90),
                chromosome = anno_simple(df_col['chromosome'], add_text=False, legend_kws={'frameon':False}, 
                                         colors= dict(zip(np.arange(1,23).astype(str),['#5daab4','#000000']*11)), 
                                         legend=True), axis=1)

cm = ClusterMapPlotter(data=data, 
                       vmax=1.02, vmin=0.98, 
                       top_annotation=top_annotation,   
                       left_annotation =right_annotation ,
                       cmap='RdBu_r',  
                       row_cluster=False, col_cluster=False, 
                        col_split = df_col['chromosome'], 
                       col_split_order = [str(i) for i in list(range(1,23))],
                       col_split_gap=0.4 
                      )

When use sns.set_style("whitegrid")
or change setting in scanpy:

import scanpy
sc.settings.verbosity = 3 
sc.settings.set_figure_params(dpi=100, dpi_save=300,
                              facecolor='white', figsize=(4,4))

An unexpected grid would appear:

When set sns.set_style("whitegrid") or in default, the heatmap looks good:

Describe the bug
I have a matrix of size 157 rows (genes) × 1023 columns (cells). When I create a clustermap plot, I get these thick white lines/ artifacts appearing in the image.

Expected behavior

When I reduce the size of my matrix 59 rows (genes) × 1023 columns (cells), the clustermap outputs as it should.

What can I do about this?

Desktop (please complete the following information):
-Linux Mint 20.1

• Chrome
• Jupyter Notebook

Additional context
Plots generated following this tutorial with own data

Hello, thanks for developing this useful tool. I encountered while using the composite() function in PyComplexHeatmap: TypeError: composite() got an unexpected keyword argument 'height_ratios'.

I think this is the version issue that was recently added, but not completely updated in the package.
Could you provide an update to the PyComplexHeatmap package to be able to address this problem?

Thank you in advance!

Dear PyComplexHeatmap developers,

Thank you for the great package. I can foresee that it will speedup our data analysis.

I encountered a value error when running the example from the readme: 'lightgray' is not a valid color value.

The error can be tracked back to line 13 in the colors.py file.

Since I seem to be the only one having this issue, I thought about package compatibility issues.
Could it be that my matplotlib's version is not supported?
I'm using matplotlib 3.8.0

Thank you for checking this issue!

Here is the complete error message: