benjamindoran/unidip

Python port of the Unidip clustering algorithm from: http://www.kdd.org/kdd2016/subtopic/view/skinny-dip-clustering-in-a-sea-of-noise

unidip's Issues

index 0 is out of bounds for axis 0 with size 0

When I run the code on a bigger amount of data, it either gives an error or the results are not so adequate. Any idea what I can do to make it work?

RuntimeWarning: divide by zero encountered in true_divide & invalid value encountered in multiply

Hi @BenjaminDoran ,
I find unidip very helpful in identifying unimodal distributions. Thanks for the development of this package!

Regarding the warning I get, this happens when I run dip.diptst

/python3.6/site-packages/unidip/dip.py:27: RuntimeWarning: divide by zero encountered in true_divide
  slopes = (work_cdf[1:] - work_cdf[0]) / distances
/python3.6/site-packages/unidip/dip.py:30: RuntimeWarning: invalid value encountered in multiply
  gcm.extend(work_cdf[0] + distances[:minslope_idx] * minslope)

After the warning, the process stalls for a very long time.
Looking at the values, I couldn't pinpoint the reason why this is happening. Any suggestions?
I wonder how I could avoid this warning.

UniDip returns empty array

Running it with this list returns an empty array, is this expected?
Shouldn't it return at least one result?

[-1.7321488631490713, -1.7321488631490713, -1.4811330900800335, -1.4811330900800335, -1.3022212430204045, -1.2717876122825453, -1.211729705204752, -1.211729705204752, -1.211729705204752, -1.0643373712891142, -1.0643373712891142, -1.0068300836569808, -1.0068300836569808, -1.0068300836569808, -1.0068300836569808, -0.93488438396035, -0.9240098946317632, -0.9210480479452614, -0.9210480479452614, -0.9169604252707326, -0.910481737919528, -0.910481737919528, -0.910481737919528, -0.8884333993387736, -0.8884333993387736, -0.8774734971225411, -0.8662954695554639, -0.8528953736209124, -0.8347337812492381, -0.8347337812492381, -0.8254311371686884, -0.8254311371686884, -0.8254311371686884, -0.8254311371686884, -0.8254311371686884, -0.8254311371686884, -0.8254311371686884, -0.8254311371686884, -0.8186966057281344, -0.7822676720935519, -0.7732098730652197, -0.7732098730652197, -0.7732098730652197, -0.7732098730652197, -0.7732098730652197, -0.7732098730652197, -0.7732098730652197, -0.7604371692027225, -0.7604371692027225, -0.7184651234303354, -0.7184651234303354, -0.7184651234303354, -0.7184651234303354, -0.7184651234303354, -0.7184651234303354, -0.7184651234303354, -0.7184651234303354, -0.7184651234303354, -0.7184651234303354, -0.6562230498767323, -0.6562230498767323, -0.6562230498767323, -0.6562230498767323, -0.6562230498767323, -0.6562230498767323, -0.6562230498767323, -0.6562230498767323, -0.6429924962522777, -0.6429924962522777, -0.6429924962522777, -0.6429924962522777, -0.6429924962522777, -0.6429924962522777, -0.6069943256197385, -0.6069943256197385, -0.6069943256197385, -0.6069943256197385, -0.6069943256197385, -0.6069943256197385, -0.5954041171436801, -0.5431426862261441, -0.5431426862261441, -0.5142420706343538, -0.5142420706343538, -0.5142420706343538, -0.5142420706343538, -0.5142420706343538, -0.5142420706343538, -0.5142420706343538, -0.5142420706343538, -0.4758761151656534, -0.4758761151656534, -0.4597064652077143, -0.4597064652077143, -0.4597064652077143, -0.4480050230671342, -0.4480050230671342, -0.2140861382454502, -0.2140861382454502, -0.2140861382454502, -0.2140861382454502, -0.1588138036471789, -0.1588138036471789, -0.036075132086434536, 0.14230167504323354, 0.5357475959080094, 0.5357475959080094, 0.5716170975364421, 0.5763855126721228, 0.5765579831608809, 0.5765579831608809, 0.5765579831608809, 0.588359377758227, 0.588359377758227, 0.588359377758227, 0.5958301926543572, 0.5958301926543572, 0.5958301926543572, 0.5958301926543572, 0.5958301926543572, 0.5958301926543572, 0.6212950129383263, 0.6212950129383263, 0.6212950129383263, 0.6212950129383263, 0.6212950129383263, 0.6325814867346722, 0.6325814867346722, 0.6325814867346722, 0.6325814867346722, 0.6325814867346722, 0.6389869403409391, 0.6389869403409391, 0.6584198203060039, 0.6584198203060039, 0.6584198203060039, 0.6584198203060039, 0.6584198203060039, 0.6584198203060039, 0.6958291014778762, 0.6958291014778762, 0.6990612564591703, 0.7119598760986781, 0.7119598760986781, 0.7211152863447583, 0.7211152863447583, 0.7211152863447583, 0.7211152863447583, 0.7211152863447583, 0.7211152863447583, 0.7211152863447583, 0.7211152863447583, 0.7211152863447583, 0.7211152863447583, 0.7211152863447583, 0.7211152863447583, 0.7211152863447583, 0.725389613191505, 0.725389613191505, 0.725389613191505, 0.725389613191505, 0.725389613191505, 0.725389613191505, 0.814682710085783, 0.814682710085783, 0.814682710085783, 0.8275420646774649, 0.8275420646774649, 0.8275420646774649, 0.8275420646774649, 0.9046055225669343, 0.9046055225669343, 0.9046055225669343, 0.9054035860830063, 0.9054035860830063, 0.9054035860830063, 0.9459602397628304, 0.9459602397628304, 0.9496206931864464, 0.9496206931864464, 0.9496206931864464, 0.9496206931864464, 0.9504804742173372, 0.9504804742173372, 0.9978027326125107, 0.9978027326125107, 0.9978027326125107, 1.0315973696216347, 1.0315973696216347, 1.0315973696216347, 1.0315973696216347, 1.0315973696216347, 1.0646299267140256, 1.1272302726476822, 1.1272302726476822, 1.2532571146831004, 1.2532571146831004, 1.3733262005067424, 1.62906282973043, 1.62906282973043]

no module named 'unidip'

Per your website instructions, I used 'from unidip import UniDip' and I got an error on the import statement saying there is no module named unidip. I used pip to install it like your page says, and using pip list it shows up there. Cannot figure out why it's not recognizing it.

I haven't even tried to use it yet, just importing to make sure it works.

I'm using anaconda.

Return same values when running recursively

I'm looping through alpha values to test the amount of peaks each finds, and I get the same output every time. I get different values if I run each alpha individually.

for i in [0.06,0.055,0.05,0.04]:#np.arange(0.07,0.05,-0.002):
data = np.msort(np.array(analysisuncomp).flatten())
print(i)
print(UniDip(data,alpha=i,ntrials=500).run())

Output:
0.06
[(6, 127), (135, 181), (196, 313), (370, 5683), (5685, 10540), (10545, 30064)]
0.055
[(6, 127), (135, 181), (196, 313), (370, 5683), (5685, 10540), (10545, 30064)]
0.05
[(6, 127), (135, 181), (196, 313), (370, 5683), (5685, 10540), (10545, 30064)]
0.04