I have two Dataframes, one containing my data read in from a CSV file and another that has the data grouped by all of the columns but the last and reindexed to contain a column for the count of the size of the groups.
df_k1 = pd.read_csv(filename, sep=';')
columns_for_groups = list(df_k1.columns)[:-1]
k1_grouped = df_k1.groupby(columns_for_groups).size().reset_index(name="Count")
I need to create a series such that every row(i) in the series corresponds to row(i) in my original Dataframe but the contents of the series need to be the size of the group that the row belongs to in the grouped Dataframe. I currently have this, and it works for my purposes, but I was wondering if anyone knew of a faster or more elegant solution.
size_by_row = []
for row in df_k1.itertuples():
for group in k1_grouped.itertuples():
if row[1:-1] == group[1:-1]:
size_by_row.append(group[-1])
break
group_size = pd.Series(size_by_row)
Related
I have a big dataframe its about 200k of rows and 3 columns (x, y, z). Some rows doesn't have y,z values and just have x value. I want to make a new column that first set of data with z value be 1,second one be 2,then 3, etc. Or make a multiIndex same format.
Following image shows what I mean
Like this image
I made a new column called "NO." and put zero as initial value. Then
I tried to record the index of where I want the new column get a new value. with following code
df = pd.read_fwf(path, header=None, names=['x','y','z'])
df['NO.']=0
index_NO_changed = df.index[df['z'].isnull()]
Then I loop through it and change the number:
for i in range(len(index_NO_changed)-1):
df['NO.'].iloc[index_NO_changed[i]:index_NO_changed[i+1]]=i+1
df['NO.'].iloc[index_NO_changed[-1]:]=len(index_NO_changed)
But the problem is I get a warning that "
A value is trying to be set on a copy of a slice from a DataFrame
I was wondering
Is there any better way? Is creating multiIndex instead of adding another column easier considering size of dataframe?
for row in range(1, len(df)):
try:
df_out, orthogroup, len_group = HOG_get_group_stats(df.loc[row, "HOG"])
temp_df = pd.DataFrame()
for id in range(len(df_out)):
print(" ")
temp_df = pd.concat([df, pd.DataFrame(df_out.iloc[id, :]).T], axis=1)
temp_df["HOG"] = orthogroup
temp_df["len_group"] = len_group
print(temp_df)
except:
print(row, "no")
Here I have a script that does the following:
Iterate over df and apply the HOG_get_group_stats function to the HOG column in df and then, get 3 variables as outputs. (Basically, the function creates some stats as a data frame called df_out, and extracts some information as two more columns called orthogroup, len_group)
Create an empty template called temp_df
Transpose the df_out data frame and make it one single row and then, concatenate with the df we used in the beginning as columns.
Add orthogroup, len_group columns to the end of the temp_df
Problem:
It prints out the data however, when I try to see the temp_df as a data frame it shows only a single row ( probably the last one) means that my concatenation of several data frames doesn't work.
Questions:
How can I iterate and then append a data frame as columns?
Is there an easier way to iterate over a data frame? (e.g. iterrow)
Is there a better way to transpose rows to columns in a data frame? ( e.g. pivot, melt)
Any help would be appreciated!!
You can find the sample files to df, df_out,temp_df and expected output_sample table here :
Sample_files
I have a task to compare two dataframe with same columns name but different size, we can call it previous and current. I am trying to get the difference between (previous and current) in the Quantity and Booked Columns and highlight it as yellow. The common key between the two dataframe would be the 'SN' columns
I have coded out the following
for idx, rows in df_n.iterrows():
if rows["Quantity"] == rows['Available'] + rows['Booked']:
continue
else:
rows["Quantity"] = rows["Quantity"] - rows['Available'] - rows['Booked']
df_n.loc[idx, 'Quantity'].style.applymap('background-color: yellow')
# pdb.set_trace()
if (df_o['Booked'][df_o['SN'] == rows["SN"]] != rows['Booked']).bool():
df_n.loc[idx, 'Booked'].style.apply('background-color: yellow')
I realise I have a few problems here and need some help
df_n.loc[idx, 'Quantity'] returns value instead of a dataframe type. How can I get a dataframe from one cell. Do I have to pd.DataFrame(data=df_n.loc[idx, 'Quantity'], index=idx, columns ='Quantity'). Will this create a copy or will update the reference?
How do I compare the SN of both dataframe, looking for a better way to compare. One thing I could think of is to use set index for both dataframe and when finished using them, reset them back?
My dataframe:
Previous dataframe
Current Dataframe
df_n.loc[idx, 'Quantity'] returns value instead of a dataframe type.
How can I get a dataframe from one cell. Do I have to
pd.DataFrame(data=df_n.loc[idx, 'Quantity'], index=idx, columns
='Quantity'). Will this create a copy or will update the reference?
To create a DataFrame from one cell you can try: df_n.loc[idx, ['Quantity']].to_frame().T
How do I compare the SN of both dataframe, looking for a better way to
compare. One thing I could think of is to use set index for both
dataframe and when finished using them, reset them back?
You can use df_n.merge(df_o, on='S/N') to merge dataframes and 'compare' columns.
I am selecting row by row as follows:
for i in range(num_rows):
row = df.iloc[i]
as a result I am getting a Series object where row.index.values contains names of df columns.
But I wanted instead dataframe with only one row having dataframe columns in place.
When I do row.to_frame() instead of 1x85 dataframe (1 row, 85 cols) I get 85x1 dataframe where index contains names of columns and row.columns
outputs
Int64Index([0], dtype='int64').
But all I want is just original data-frame columns with only one row. How do I do it?
Or how do I convert row.index values to row.column values and change 85x1 dimension to 1x85
You just need to adding T
row.to_frame().T
Also change your for loop with adding []
for i in range(num_rows):
row = df.iloc[[i]]
I have a CSV file whose columns are frequency counts of words, and whose rows are time periods. I want to sum for each column the total frequencies. Then I want to write to a CSV file for sums greater than or equal to 30, the column and row values, thus dropping columns whose sums are less than 30.
Just learning python and pandas. I know it is a simple question, but my knowledge is at that level. Your help is most appreciated.
I can read in the CSV file and compute the column sums.
df = pd.read_csv('data.csv')
Except of data file containing 3,874 columns and 100 rows
df.sum(axis = 0, skipna = True)
Excerpt of sums for columns
I am stuck on how to create the output file so that it looks like the original file but no longer has columns whose sums were less than 30.
I am stuck on how to write to a CSV file each row for each column whose sums are greater than or equal to 30. The layout of the output file would be the same as for the input file. The sums would not be included in the output.
Thanks very much for your help.
So, here is a link showing an excerpt of a file containing 100 rows and 3,857 columns:
It's easiest to do this in two steps:
1. Filter the DataFrame to just the columns you want to save
df_to_save = df.loc[:, (df.sum(axis=0, skipna=True) >= 30)]
.loc is for picking rows/columns based either on labels or conditions; the syntax is .loc[rows, columns], so : means "take all the rows", and then the second part is the condition on our columns - I've taken the sum you'd given in your question and set it greater than or equal to 30.
2. Save the filtered DataFrame to CSV
df_to_save.to_csv('path/to/write_file.csv', header=True, index=False)
Just put your filepath in as the first argument. header=True means the header labels from the table will be written back out to the file, and index=False means the numbered row labels Pandas automatically created when you read in the CSV won't be included in the export.
See this answer here: How to delete a column in pandas dataframe based on a condition? . Note, the solution for your question doesn't need isnull() before the sum(), as that is specific to their question for counting NaN values.