ValueError: Wrong number of items passed – Meaning and suggestions?
Question:
I am receiving the error:
ValueError: Wrong number of items passed 3, placement implies 1, and I am struggling to figure out where, and how I may begin addressing the problem.
I don’t really understand the meaning of the error; which is making it difficult for me to troubleshoot. I have also included the block of code that is triggering the error in my Jupyter Notebook.
The data is tough to attach; so I am not looking for anyone to try and re-create this error for me. I am just looking for some feedback on how I could address this error.
KeyError Traceback (most recent call last)
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandasindexesbase.py in get_loc(self, key, method, tolerance)
1944 try:
-> 1945 return self._engine.get_loc(key)
1946 except KeyError:
pandasindex.pyx in pandas.index.IndexEngine.get_loc (pandasindex.c:4154)()
pandasindex.pyx in pandas.index.IndexEngine.get_loc (pandasindex.c:4018)()
pandashashtable.pyx in pandas.hashtable.PyObjectHashTable.get_item (pandashashtable.c:12368)()
pandashashtable.pyx in pandas.hashtable.PyObjectHashTable.get_item (pandashashtable.c:12322)()
KeyError: 'predictedY'
During handling of the above exception, another exception occurred:
KeyError Traceback (most recent call last)
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandascoreinternals.py in set(self, item, value, check)
3414 try:
-> 3415 loc = self.items.get_loc(item)
3416 except KeyError:
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandasindexesbase.py in get_loc(self, key, method, tolerance)
1946 except KeyError:
-> 1947 return self._engine.get_loc(self._maybe_cast_indexer(key))
1948
pandasindex.pyx in pandas.index.IndexEngine.get_loc (pandasindex.c:4154)()
pandasindex.pyx in pandas.index.IndexEngine.get_loc (pandasindex.c:4018)()
pandashashtable.pyx in pandas.hashtable.PyObjectHashTable.get_item (pandashashtable.c:12368)()
pandashashtable.pyx in pandas.hashtable.PyObjectHashTable.get_item (pandashashtable.c:12322)()
KeyError: 'predictedY'
During handling of the above exception, another exception occurred:
ValueError Traceback (most recent call last)
<ipython-input-95-476dc59cd7fa> in <module>()
26 return gp, results
27
---> 28 gp_dailyElectricity, results_dailyElectricity = predictAll(3, 0.04, trainX_dailyElectricity, trainY_dailyElectricity, testX_dailyElectricity, testY_dailyElectricity, testSet_dailyElectricity, 'Daily Electricity')
<ipython-input-95-476dc59cd7fa> in predictAll(theta, nugget, trainX, trainY, testX, testY, testSet, title)
8
9 results = testSet.copy()
---> 10 results['predictedY'] = predictedY
11 results['sigma'] = sigma
12
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandascoreframe.py in __setitem__(self, key, value)
2355 else:
2356 # set column
-> 2357 self._set_item(key, value)
2358
2359 def _setitem_slice(self, key, value):
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandascoreframe.py in _set_item(self, key, value)
2422 self._ensure_valid_index(value)
2423 value = self._sanitize_column(key, value)
-> 2424 NDFrame._set_item(self, key, value)
2425
2426 # check if we are modifying a copy
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandascoregeneric.py in _set_item(self, key, value)
1462
1463 def _set_item(self, key, value):
-> 1464 self._data.set(key, value)
1465 self._clear_item_cache()
1466
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandascoreinternals.py in set(self, item, value, check)
3416 except KeyError:
3417 # This item wasn't present, just insert at end
-> 3418 self.insert(len(self.items), item, value)
3419 return
3420
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandascoreinternals.py in insert(self, loc, item, value, allow_duplicates)
3517
3518 block = make_block(values=value, ndim=self.ndim,
-> 3519 placement=slice(loc, loc + 1))
3520
3521 for blkno, count in _fast_count_smallints(self._blknos[loc:]):
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandascoreinternals.py in make_block(values, placement, klass, ndim, dtype, fastpath)
2516 placement=placement, dtype=dtype)
2517
-> 2518 return klass(values, ndim=ndim, fastpath=fastpath, placement=placement)
2519
2520 # TODO: flexible with index=None and/or items=None
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandascoreinternals.py in __init__(self, values, placement, ndim, fastpath)
88 raise ValueError('Wrong number of items passed %d, placement '
89 'implies %d' % (len(self.values),
---> 90 len(self.mgr_locs)))
91
92 @property
ValueError: Wrong number of items passed 3, placement implies 1
My code is as follows:
def predictAll(theta, nugget, trainX, trainY, testX, testY, testSet, title):
gp = gaussian_process.GaussianProcess(theta0=theta, nugget =nugget)
gp.fit(trainX, trainY)
predictedY, MSE = gp.predict(testX, eval_MSE = True)
sigma = np.sqrt(MSE)
results = testSet.copy()
results['predictedY'] = predictedY
results['sigma'] = sigma
print ("Train score R2:", gp.score(trainX, trainY))
print ("Test score R2:", sklearn.metrics.r2_score(testY, predictedY))
plt.figure(figsize = (9,8))
plt.scatter(testY, predictedY)
plt.plot([min(testY), max(testY)], [min(testY), max(testY)], 'r')
plt.xlim([min(testY), max(testY)])
plt.ylim([min(testY), max(testY)])
plt.title('Predicted vs. observed: ' + title)
plt.xlabel('Observed')
plt.ylabel('Predicted')
plt.show()
return gp, results
gp_dailyElectricity, results_dailyElectricity = predictAll(3, 0.04, trainX_dailyElectricity, trainY_dailyElectricity, testX_dailyElectricity, testY_dailyElectricity, testSet_dailyElectricity, 'Daily Electricity')
Answers:
In general, the error ValueError: Wrong number of items passed 3, placement implies 1 suggests that you are attempting to put too many pigeons in too few pigeonholes. In this case, the value on the right of the equation
results['predictedY'] = predictedY
is trying to put 3 “things” into a container that allows only one. Because the left side is a dataframe column, and can accept multiple items on that (column) dimension, you should see that there are too many items on another dimension.
Here, it appears you are using sklearn for modeling, which is where gaussian_process.GaussianProcess() is coming from (I’m guessing, but correct me and revise the question if this is wrong).
Now, you generate predicted values for y here:
predictedY, MSE = gp.predict(testX, eval_MSE = True)
However, as we can see from the documentation for GaussianProcess, predict() returns two items. The first is y, which is array-like (emphasis mine). That means that it can have more than one dimension, or, to be concrete for thick headed people like me, it can have more than one column — see that it can return (n_samples, n_targets) which, depending on testX, could be (1000, 3) (just to pick numbers). Thus, your predictedY might have 3 columns.
If so, when you try to put something with three “columns” into a single dataframe column, you are passing 3 items where only 1 would fit.
Not sure if this is relevant to your question but it might be relevant to someone else in the future: I had a similar error. Turned out that the df was empty (had zero rows) and that is what was causing the error in my command.
for i in range(100):
try:
#Your code here
break
except:
continue
This one worked for me.
So ValueError: The wrong number of items passed 3, placement implies 1 occurs when you’re passing to many arguments but method supports only a few. for example –
df['First_Name', 'Last_Name'] = df['Full_col'].str.split(' ', expand = True)
In the above code, I’m trying to split Full_col into two sub-columns names as -First_Name & Last_Name, so here I’ll get the error because instead list of columns the columns I’m passing only a single argument.
So to avoid this – use another sub-list
df[['First_Name', 'Last_Name']] = df['Full_col'].str.split(' ', expand = True)
Another cause of this error is when you apply a function on a DataFrame where there are two columns with the same name.
Just adding this as an answer: nesting methods and misplacing closed brackets will also throw this error, ex:
march15_totals= march15_t.assign(sum_march15_t=march15_t[{"2021-03-15","2021-03-16","2021-03-17","2021-03-18","2021-03-19","2021-03-20","2021-03-21"}]).sum(axis=1)
Versus the (correct) version:
march15_totals= march15_t.assign(sum_march15_t=march15_t[{"2021-03-15","2021-03-16","2021-03-17","2021-03-18","2021-03-19","2021-03-20","2021-03-21"}].sum(axis=1))
This is probably common sense to most of you but I was quite puzzled until I realized my mistake.
Starting with pandas 1.3.x it’s not allowed to fill objects (e.g. like an eagertensor from an embedding) into columns.
https://github.com/pandas-dev/pandas/blame/master/pandas/core/internals/blocks.py
I got this error when I was trying to convert a one-column dataframe, df, into a Series, pd.Series(df).
I resolved this with
pd.Series(df.values.flatten())
The problem was that the values in the dataframe were lists:
my_col
0 ['a']
1 ['b']
2 ['c']
3 ['d']
When I was printing the dataframe it wasn’t showing the brackets which made it hard to track down.
I am receiving the error:
ValueError: Wrong number of items passed 3, placement implies 1, and I am struggling to figure out where, and how I may begin addressing the problem.
I don’t really understand the meaning of the error; which is making it difficult for me to troubleshoot. I have also included the block of code that is triggering the error in my Jupyter Notebook.
The data is tough to attach; so I am not looking for anyone to try and re-create this error for me. I am just looking for some feedback on how I could address this error.
KeyError Traceback (most recent call last)
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandasindexesbase.py in get_loc(self, key, method, tolerance)
1944 try:
-> 1945 return self._engine.get_loc(key)
1946 except KeyError:
pandasindex.pyx in pandas.index.IndexEngine.get_loc (pandasindex.c:4154)()
pandasindex.pyx in pandas.index.IndexEngine.get_loc (pandasindex.c:4018)()
pandashashtable.pyx in pandas.hashtable.PyObjectHashTable.get_item (pandashashtable.c:12368)()
pandashashtable.pyx in pandas.hashtable.PyObjectHashTable.get_item (pandashashtable.c:12322)()
KeyError: 'predictedY'
During handling of the above exception, another exception occurred:
KeyError Traceback (most recent call last)
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandascoreinternals.py in set(self, item, value, check)
3414 try:
-> 3415 loc = self.items.get_loc(item)
3416 except KeyError:
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandasindexesbase.py in get_loc(self, key, method, tolerance)
1946 except KeyError:
-> 1947 return self._engine.get_loc(self._maybe_cast_indexer(key))
1948
pandasindex.pyx in pandas.index.IndexEngine.get_loc (pandasindex.c:4154)()
pandasindex.pyx in pandas.index.IndexEngine.get_loc (pandasindex.c:4018)()
pandashashtable.pyx in pandas.hashtable.PyObjectHashTable.get_item (pandashashtable.c:12368)()
pandashashtable.pyx in pandas.hashtable.PyObjectHashTable.get_item (pandashashtable.c:12322)()
KeyError: 'predictedY'
During handling of the above exception, another exception occurred:
ValueError Traceback (most recent call last)
<ipython-input-95-476dc59cd7fa> in <module>()
26 return gp, results
27
---> 28 gp_dailyElectricity, results_dailyElectricity = predictAll(3, 0.04, trainX_dailyElectricity, trainY_dailyElectricity, testX_dailyElectricity, testY_dailyElectricity, testSet_dailyElectricity, 'Daily Electricity')
<ipython-input-95-476dc59cd7fa> in predictAll(theta, nugget, trainX, trainY, testX, testY, testSet, title)
8
9 results = testSet.copy()
---> 10 results['predictedY'] = predictedY
11 results['sigma'] = sigma
12
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandascoreframe.py in __setitem__(self, key, value)
2355 else:
2356 # set column
-> 2357 self._set_item(key, value)
2358
2359 def _setitem_slice(self, key, value):
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandascoreframe.py in _set_item(self, key, value)
2422 self._ensure_valid_index(value)
2423 value = self._sanitize_column(key, value)
-> 2424 NDFrame._set_item(self, key, value)
2425
2426 # check if we are modifying a copy
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandascoregeneric.py in _set_item(self, key, value)
1462
1463 def _set_item(self, key, value):
-> 1464 self._data.set(key, value)
1465 self._clear_item_cache()
1466
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandascoreinternals.py in set(self, item, value, check)
3416 except KeyError:
3417 # This item wasn't present, just insert at end
-> 3418 self.insert(len(self.items), item, value)
3419 return
3420
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandascoreinternals.py in insert(self, loc, item, value, allow_duplicates)
3517
3518 block = make_block(values=value, ndim=self.ndim,
-> 3519 placement=slice(loc, loc + 1))
3520
3521 for blkno, count in _fast_count_smallints(self._blknos[loc:]):
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandascoreinternals.py in make_block(values, placement, klass, ndim, dtype, fastpath)
2516 placement=placement, dtype=dtype)
2517
-> 2518 return klass(values, ndim=ndim, fastpath=fastpath, placement=placement)
2519
2520 # TODO: flexible with index=None and/or items=None
C:Usersbrennn1AppDataLocalContinuumAnaconda3libsite-packagespandascoreinternals.py in __init__(self, values, placement, ndim, fastpath)
88 raise ValueError('Wrong number of items passed %d, placement '
89 'implies %d' % (len(self.values),
---> 90 len(self.mgr_locs)))
91
92 @property
ValueError: Wrong number of items passed 3, placement implies 1
My code is as follows:
def predictAll(theta, nugget, trainX, trainY, testX, testY, testSet, title):
gp = gaussian_process.GaussianProcess(theta0=theta, nugget =nugget)
gp.fit(trainX, trainY)
predictedY, MSE = gp.predict(testX, eval_MSE = True)
sigma = np.sqrt(MSE)
results = testSet.copy()
results['predictedY'] = predictedY
results['sigma'] = sigma
print ("Train score R2:", gp.score(trainX, trainY))
print ("Test score R2:", sklearn.metrics.r2_score(testY, predictedY))
plt.figure(figsize = (9,8))
plt.scatter(testY, predictedY)
plt.plot([min(testY), max(testY)], [min(testY), max(testY)], 'r')
plt.xlim([min(testY), max(testY)])
plt.ylim([min(testY), max(testY)])
plt.title('Predicted vs. observed: ' + title)
plt.xlabel('Observed')
plt.ylabel('Predicted')
plt.show()
return gp, results
gp_dailyElectricity, results_dailyElectricity = predictAll(3, 0.04, trainX_dailyElectricity, trainY_dailyElectricity, testX_dailyElectricity, testY_dailyElectricity, testSet_dailyElectricity, 'Daily Electricity')
In general, the error ValueError: Wrong number of items passed 3, placement implies 1 suggests that you are attempting to put too many pigeons in too few pigeonholes. In this case, the value on the right of the equation
results['predictedY'] = predictedY
is trying to put 3 “things” into a container that allows only one. Because the left side is a dataframe column, and can accept multiple items on that (column) dimension, you should see that there are too many items on another dimension.
Here, it appears you are using sklearn for modeling, which is where gaussian_process.GaussianProcess() is coming from (I’m guessing, but correct me and revise the question if this is wrong).
Now, you generate predicted values for y here:
predictedY, MSE = gp.predict(testX, eval_MSE = True)
However, as we can see from the documentation for GaussianProcess, predict() returns two items. The first is y, which is array-like (emphasis mine). That means that it can have more than one dimension, or, to be concrete for thick headed people like me, it can have more than one column — see that it can return (n_samples, n_targets) which, depending on testX, could be (1000, 3) (just to pick numbers). Thus, your predictedY might have 3 columns.
If so, when you try to put something with three “columns” into a single dataframe column, you are passing 3 items where only 1 would fit.
Not sure if this is relevant to your question but it might be relevant to someone else in the future: I had a similar error. Turned out that the df was empty (had zero rows) and that is what was causing the error in my command.
for i in range(100):
try:
#Your code here
break
except:
continue
This one worked for me.
So ValueError: The wrong number of items passed 3, placement implies 1 occurs when you’re passing to many arguments but method supports only a few. for example –
df['First_Name', 'Last_Name'] = df['Full_col'].str.split(' ', expand = True)
In the above code, I’m trying to split Full_col into two sub-columns names as -First_Name & Last_Name, so here I’ll get the error because instead list of columns the columns I’m passing only a single argument.
So to avoid this – use another sub-list
df[['First_Name', 'Last_Name']] = df['Full_col'].str.split(' ', expand = True)
Another cause of this error is when you apply a function on a DataFrame where there are two columns with the same name.
Just adding this as an answer: nesting methods and misplacing closed brackets will also throw this error, ex:
march15_totals= march15_t.assign(sum_march15_t=march15_t[{"2021-03-15","2021-03-16","2021-03-17","2021-03-18","2021-03-19","2021-03-20","2021-03-21"}]).sum(axis=1)
Versus the (correct) version:
march15_totals= march15_t.assign(sum_march15_t=march15_t[{"2021-03-15","2021-03-16","2021-03-17","2021-03-18","2021-03-19","2021-03-20","2021-03-21"}].sum(axis=1))
This is probably common sense to most of you but I was quite puzzled until I realized my mistake.
Starting with pandas 1.3.x it’s not allowed to fill objects (e.g. like an eagertensor from an embedding) into columns.
https://github.com/pandas-dev/pandas/blame/master/pandas/core/internals/blocks.py
I got this error when I was trying to convert a one-column dataframe, df, into a Series, pd.Series(df).
I resolved this with
pd.Series(df.values.flatten())
The problem was that the values in the dataframe were lists:
my_col
0 ['a']
1 ['b']
2 ['c']
3 ['d']
When I was printing the dataframe it wasn’t showing the brackets which made it hard to track down.