What should fill_between do if it gets a greater than 1D array?

I think the change we should make here is to clarify that the inputs should be 1D and raise a better exception.

np.squeeze might be a better option that the above slicing.

The problem originally came out when reading data from a file. The vectors were one dimenional but second dimension was fixed to 1 rather than empty. I think that case should be handled even when where=None.

If data.ndim != 1 it is not 1D data (in the most pedantic way).

If we start to take (N, 1) as input, then we should probably take (1, N) as valid input as well. We then would have to decide how to deal with the user providing a mix if (N,), (N, 1), and (1, N). Should we just squeeze them or is that an error? I suspect that either answer will be wrong in some contexts which is why I think it is better if mpl requires 1D inputs (that is np.array(v).ndim == 1) and leave the decisions about how to go from the application data -> what mpl needs to the application developers.

Ok. I see your point. I am ok with using squeeze. Maybe then just change the error message. I was initally not aware of where parameter. The confusing thing was that all arguments I was giving explictly had agreeing dimensions.

The exact work here:

• validate the inputs to fill_between and fill_betweenx are 1D (because we can also take in lists, numpy arrays, and pd.Series this should probably be cbook.is_scaler(cbook.safe_first_element(inp)) to check that it is a) an iterable b) the things that come out of iteration look like scalars).
• raise a clear error if not

Can i try this , if it is still open ?

@gauthamzz Please do! I do not think anyone is currently working on this.

Can you tell me the names of the associated files ?

Both of those methods are in lib/matplotlib/axes/_axes.py

I am trying to run control chart module got below error . can you provide solution for this below error ?
ValueError Traceback (most recent call last)
in
7 cc_data = np.array(ppmdata[x_names_significant_subprocess].values.ravel().reshape(-1, 1))
8 message_cc = control_chart(x_names_significant_subprocess, directory, ccdata, target_mean, sigma,
----> 9 sub_group_size)

F:\Python Scripts\ppmBOTv07\bot\api\ppmutils.py in control_chart(MS_SP_X, directory, cc_data, target_mean, sigma, sub_group_size)
1198 chart1 = spc(target=target_mean, data=cc_data, stdev=sigma) + xmr() + mr() + rules()
1199 filename = 'CC_' + Chart_Type + datetime.datetime.now().strftime("_%Y-%m-%d-%H%M%S") + '.jpg'
-> 1200 chart1.save(os.path.join(directory,filename))
1201 plt.close()
1202 message = "Chart is saved in the Charts directory."

F:\Python Scripts\ppmBOTv07\bot\spc\spc.py in save(self, filename, **kwargs)
101 """
102 if len(self.summary) == 0:
--> 103 self.make()
104
105 nv, normp = ntest(self.datas)

F:\Python Scripts\ppmBOTv07\bot\spc\spc.py in make(self, **kwargs)
154 utl = self.t + 3 * self.sigma
155 ltl = max(self.t - 3 * self.sigma, 0)
--> 156 PlotCharts(ax, values, self.t, center, lcl, ucl, ltl, utl, title)
157
158 summary['name'] = title

F:\Python Scripts\ppmBOTv07\bot\spc\results.py in init(self, ax, values, target, center, lcl, ucl, ltl, utl, title)
27 super(PlotCharts, self).init()
28 plt.style.use('ggplot')
---> 29 self.plot_chart(ax, values, target, center, lcl, ucl, ltl, utl, title)
30
31 def plot_chart(self, ax, values, target, center, lcl, ucl, ltl, utl, title, newvalues=None):

F:\Python Scripts\ppmBOTv07\bot\spc\results.py in plot_chart(self, ax, values, target, center, lcl, ucl, ltl, utl, title, newvalues)
62
63 else:
---> 64 ax.fill_between([-0.3, num], [lcl, lcl], [ucl, ucl], facecolor='green', alpha=0.4)
65 if lcl == ltl:
66 ax.yaxis.set_ticks([lcl, target, center, ucl, utl])

F:\Anaconda\lib\site-packages\matplotlib_init_.py in inner(ax, data, *args, **kwargs)
1808 "the Matplotlib list!)" % (label_namer, func.name),
1809 RuntimeWarning, stacklevel=2)
-> 1810 return func(ax, *args, **kwargs)
1811
1812 inner.doc = _add_data_doc(inner.doc,

F:\Anaconda\lib\site-packages\matplotlib\axes_axes.py in fill_between(self, x, y1, y2, where, interpolate, step, **kwargs)
5052 if array.ndim > 1:
5053 raise ValueError('Input passed into argument "%r"' % name +
-> 5054 'is not 1-dimensional.')
5055
5056 if where is None:

ValueError: Input passed into argument "'y1'"is not 1-dimensional.

What the error message says: y1 must be a 1-dimensional array. You've passed in something else.

F:\Anaconda\lib\site-packages\matplotlib\axes_axes.py in fill_between(self, x, y1, y2, where, interpolate, step, **kwargs)
5052 if array.ndim > 1:
5053 raise ValueError('Input passed into argument "%r"' % name +
-> 5054 'is not 1-dimensional.')

this was working 2 weeks back. After I installed OS due to some reason, installed Anaconda python v3.7 this error I got

Please check the shape of y1 in your code and make it 1D.

I'm sorry this is popping up for you. However, this change just enforces what was implicitly assumed and explicitly documented on the data already. The former tolerance towards multi-dimensional arrays may or may not have worked properly depending on the actual data. And we don't want to risk wrong plots due to sloppy data verification.