drazenz / heatmaps Goto Github PK

Hi,

Is it possible to add the actual correlation score to the heatmap, similar to seaborn's annotate?

Hi,
I can't seem to plot a corrplot. Here's what I'm running:

>>> import pandas as pd
>>> from sklearn import datasets
>>> from heatmap import corrplot
>>> iris = datasets.load_iris()
>>> X = pd.DataFrame(data=iris.data, columns=iris.feature_names)
>>> y = pd.DataFrame(data=iris.target, columns=['target'])
>>> df = pd.concat([X, y], axis=1)
>>> corrplot(df.corr())

I get:

Traceback (most recent call last):
  File "/Users/shaked.z/myenv/lib/python3.6/site-packages/numpy/core/function_base.py", line 117, in linspace
    num = operator.index(num)
TypeError: 'float' object cannot be interpreted as an integer

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/Users/shaked.z/myenv/lib/python3.6/site-packages/heatmap/heatmap.py", line 129, in corrplot
    palette=sns.diverging_palette(20, 220, n=256),
  File "/Users/shaked.z/myenv/lib/python3.6/site-packages/seaborn/palettes.py", line 694, in diverging_palette
    neg = palfunc((h_neg, s, l), 128 - (sep / 2), reverse=True, input="husl")
  File "/Users/shaked.z/myenv/lib/python3.6/site-packages/seaborn/palettes.py", line 591, in light_palette
    return blend_palette(colors, n_colors, as_cmap)
  File "/Users/shaked.z/myenv/lib/python3.6/site-packages/seaborn/palettes.py", line 727, in blend_palette
    pal = _ColorPalette(pal(np.linspace(0, 1, n_colors)))
  File "<__array_function__ internals>", line 6, in linspace
  File "/Users/shaked.z/myenv/lib/python3.6/site-packages/numpy/core/function_base.py", line 121, in linspace
    .format(type(num)))
TypeError: object of type <class 'float'> cannot be safely interpreted as an integer.

Installed heatmapz 0.0.4 using pip. Thanks!

This simple thing can allow to us to add our own customizations. The most immediate one for me being mask. just add similar code from the heatmap function. also the dictionary.update function might allow me to add my own palette directly from kwargs too.

I have downloaded the code and edited a bit so that the functions will return both the axes (the one where the scatterplot is drawn and the one where the sidebar is drawn) so that users can further customize these plots for more aesthetics.

Your code just defines functions, has no main part.
Can you provide examples for how to actually use your code to produce graphics?

Very useful implementation - Please add a license for the community!

import pandas as pd
from heatmap import corrplot

tdf = pd.DataFrame([[1, -1, 3], [0, 3, 5]], columns=pd.Series(['A', 'B', 'C'], name='COLUMNS'))
test_corr = tdf.corr()
print(test_corr)
test_corr.index.name = 'index'
corrplot(test_corr)

Expected: expect this to run

Actual:
KeyError: 'index'

problem is this hardcoding:
~\Anaconda3\envs\cs\lib\site-packages\heatmap\heatmap.py in corrplot(data, size_scale, marker)
122
123 def corrplot(data, size_scale=500, marker='s'):
--> 124 corr = pd.melt(data.reset_index(), id_vars='index').replace(np.nan, 0)
125 corr.columns = ['x', 'y', 'value']
126 heatmap(

Workaround:
do test_corr.index.name = 'index' before the corrplot call

which gives the expected result:

I am running your latest code in python 2.7.16.
I get a run time error

File "/Users/tom/Desktop/Labcyte/data/heatmap.py", line 110, in heatmap
linewidth=0
TypeError: barh() takes at least 3 arguments (4 given)

My call looks like this.
corr_df = df
print corr_df[col_numer].corr()
plt.figure(figsize=(7, 7))
corrplot(corr_df[col_numer].corr().replace(np.nan, 0), size_scale=500)

What's wrong?

This approach looks really great, thank you. I haven't managed to apply formatting for fonts etc. - is it possible? I tried using Matplotlib's RC parameters, but they have no effect. Is there a way to do it?

Hey @dylan-profiler!

Thanks for this library, do you any idea on how to add the marker size to the legend/plot?

Thanks in advance!

def corrplot(data, size_scale=500, marker='s'):
corr = pd.melt(data.reset_index(), id_vars='index').replace(np.nan, 0)
corr.columns = ['x', 'y', 'value']
heatmap(
corr['x'], corr['y'],
color=corr['value'], color_range=[-1, 1],
palette=sns.diverging_palette(20, 220, n=256),
size=corr['value'].abs(), size_range=[0,1],
marker=marker,
x_order=data.columns,
y_order=data.columns[::-1],
size_scale=size_scale
)

This is the function which leads to error.
However, changing to x_order=data.index can fix the bug