Pandas Keyerror when specifying loc row labels using a loop

Question

I have the following dataframe Positive Samples Dataframe loaded from a txt file using pandas

This positive samples dataframe has a column called Gene Set, which is basically a list of genes. When I run postive_samples["Gene Set"] I get the following output

['YAL004W', 'YLL024C'] ['YAL005C', 'YLL024C'] ['YAL005C', 'YMR006C'] ['YAL005C', 'YOL090W'] ['YAL009W', 'YBR074W'] ['YAL009W', 'YER162C'] ['YAL009W', 'YHL024W'] ['YAL009W', 'YJL187C'] ['YAL009W', 'YKR003W']

I also have another dataframe called new_expression_df New Expression Dataframe, which has positive_samples["Gene Set"] column as its index.

So What i am trying to do is get the values which are stored in postive_samples["Gene Set"] as they are and locate them using loc in the new_expression_df index using a loop.

samples_column_list= ["GSM144760","GSM144761","GSM144762","GSM144763","GSM144764"]

for gene_class_column in postive_samples[['Gene Set']]:
  #Select column contents by column name using [] operator
    geneSeriesObj = postive_samples[gene_class_column]
    gene_pairs = geneSeriesObj.values

#get gene pairs and locate their expression in the given samples    
for gene_pair in gene_pairs:
  new_expression_df.loc[gene_pair,samples_column_list]

I am getting a key Error at the beginning of the iteration when I try to do this using a loop, ideally I want to get each gene set as a list, locate its values in another data frame using its index.

But, when I plug in each set like below without using a loop it works just fine for the same values I get a Key Error for, so what am I doing wrong here?

new_expression_df.loc[['YAL002W','YBL001C'],samples_column_list]

I want to put the row argument of the loc function dynamically from another data frame column values which are stored in a list.

Shradha · Accepted Answer · 2020-09-25 12:21:19Z

1

Can you print gene_pairs just before the second for loop?

edited Sep 25, 2020 at 12:21

answered Sep 25, 2020 at 12:16

Shradha

2,4721 gold badge16 silver badges27 bronze badges

Sign up to request clarification or add additional context in comments.

5 Comments

Morena Over a year ago

when I print gene_pairs I get the same output highlighted above, like below on each iteration. So each iteration outputs one list ['YAL004W', 'YLL024C'] ['YAL005C', 'YLL024C'] ['YAL005C', 'YMR006C'] ['YAL005C', 'YOL090W'] ['YAL009W', 'YBR074W'] ['YAL009W', 'YER162C'] ['YAL009W', 'YHL024W'] ['YAL009W', 'YJL187C'] ['YAL009W', 'YKR003W']

Shradha Over a year ago

Can you tell me the output of the following- for gene_pair in gene_pairs: print(gene_pair) new_expression_df.loc[gene_pair,samples_column_list]

Morena Over a year ago

what that does is, it prints each pair. ['YAL004W', 'YLL024C'] is one output of print(gene_pair). A gene pair is the 2 elements of the array/list I gave above. So if I put it in a loop it outputs all the other pairs like I commented above but what I want to do is, put each output iteration in the loc function and execute it without getting the key error

Shradha Over a year ago

Try using .at instead of .loc. Also can you pass the value to print method? for gene_pair in gene_pairs: print(new_expression_df.loc[gene_pair,samples_column_list])

Morena Over a year ago

I think .loc is better compared to .at, since I am trying to return values based on more than one row and column, but I gave it a shot though and its giving me a type error if I use .at because I am giving it a list The output is still the same, I get the key error even when I wrap that in a print statement. What I am wondering is how is it having a key error when using a loop but not having any errors when I plug all the values manually into the .loc function? It does not make any sense 🤷‍♂️. I am literally plugging in the same values with or without the loop😒

Collectives™ on Stack Overflow

Pandas Keyerror when specifying loc row labels using a loop

1 Answer 1

5 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

5 Comments

Your Answer

Sign up or log in

Post as a guest

Related